Aerogel Synthesis and Application

A thesis submitted in partial fulfillment of the requirements for the degree of

Bachelor of Arts in Physics

Pomona College

By Daniel Sedlacek

Advised by:

Dr. David Tanenbaum

April 24, 2009

Abstract

Aerogel has become a material of interest to scientists in recent decades due to its

unique physical properties that give it the potential to improve technologies in a variety of

fields. In particular, aerogels offer the lowest densities and the lowest thermal conductivities

of any known solid. Silica-based aerogels were first synthesized in 1931 by Steven Kistler.

Since then, many other types of aerogels have been created, including carbon-based and clay-

based aerogels. Silica-based aerogels are the simplest and most widely studied type of

aerogel, with new uses and applications arising every day. As such, silica-based aerogels

offer a unique platform on which to base further research in the small liberal-arts college

setting.

This project examined the process of synthesizing low-density silica-based aerogels.

The goal was to create a reliable, non-toxic method, using inexpensive materials and

equipment already owned by the Pomona College Physics Department. The science involved

was built off of results achieved by the author in Physics 174: Contemporary Experimental

Physics class at Pomona College in the spring of 2008. The recipe used to synthesize the

aerogels was found on the Lawrence Berkely National Laboratory website [9]. Additions to

the recipe were implemented as suggested on William Wood’s website [24]. The recipe

incorporated the most recent technique in aerogel manufacturing, replacing

tetramethylorthosilicate (TMOS) with tetraethylorthosilicate (TEOS). Both of the chemicals

are precondensed silica precursors, with TEOS being the less toxic of the two. The gel molds

were in the shape of cylinders so that both thin and thick disks could be created depending on

desired speed of production, and desired further experimentation. Specific drying methods

were based on Tousimis Samdri-PVT-3D critical point dryer manual and suggested

techniques from Dr. Tousimis [27]. The scanning electron microscope was then used by Dr.

2

David Tanenbaum to examine the nanoscale structures of a sample achieved from this project

as well as a commercial sample owned by the Pomona College Physics Department. Further

examination was carried out with the department’s Fluke thermal imagining camera.

3

Table of Contents Chapter 1: Introduction and Background.......................................................................5

1.1 Background 1.2 Motivation 1.3 Methods

Chapter 2: Theory............................................................................................................12 2.1 Chemical Theory

2.2 Methods Theory 2.3 Physical Theory

Chapter 3: Previous Research........................................................................................17 Chapter 4: Experimental Setup and Process.................................................................23 4.1 Materials 4.2 The Gel Stage 4.3 The Mold Stage 4.4 The Bath Stage 4.5 The Drying Stage Chapter 5: Results and Analysis.....................................................................................28 Chapter 6: Conclusions and Future Work…................................................................37 6.1: Conclusions 6.2: Future Work 7: Acknowledgements......................................................................................................41 8: Bibliography.................................................................................................................42 9: Appendix I: Images of Equipment.............................................................................46 10: Appendix II: More Sample Images..........................................................................49 11: Appendix III: Student Manual for Synthesis of Simple SiO2 Aerogels................52

4

Chapter 1: Introduction

1.1: Background In 1931, Steven Kistler made a bet with a colleague that he could prove a wet gel

contained a solid matrix the same size and shape of the gel. To do this he began with a

gel and extracted the liquid, leaving a low-density solid behind. Using an autoclave to

drive the liquid past its critical point he was able to conquer the obstacle of surface

tension which would otherwise rip apart the internal solid structure of the gel. His

successful wager produced the first silica-based aerogel. For half a century this curious

material went relatively unnoticed, due to the notorious difficulties and safety issues

involved in its creation.

In the early years, fabricating aerogels meant sending alcohol to volatile pressures

and temperatures in order for it to reach its supercritical point and allow for the

supercritical-extraction of the gel. Then, in the 1980’s, interest was renewed when a

French scientist, attempting to improve the fabrication process for the French

government, developed a process which used less-toxic materials. He switched out

methyl alcohol and tetramethylorthosilicate (TMOS) for the safer pairing of ethyl alcohol

and tetraethylorthosilicate (TEOS). The next breakthrough came in the early 1990’s when

liquid carbon dioxide replaced the ethyl alcohol involved in the gel before the sample was

taken through the supercritical process. This allowed scientists to bypass the dangerous

pressures and temperatures needed to send the pure ethanol past its supercritical point.

Liquid carbon dioxide has the relatively mundane requirements of 305 K and 1050 psi to

be brought to its supercritical point (Fig.1.1).

5

Fig. 1.1 Above: CO2 Phase Diagram [29] Due to these and other breakthroughs, aerogels saw a surge in popularity in the

1990’s. Scientists found aerogel to be a new class of material altogether, with untapped

potential due to its varied and unique properties [19]. Making up for its brittleness,

aerogel holds the record as the solid with the lowest density the lowest thermal

conductivity. The lowest density evacuated aerogel ever produced had a density of 1.0

milligrams per cubic centimeter – so light that the sample could float in air (1.2

milligrams per cubic centimeter). Though many materials can be used to form the solid

lattice structure of an aerogel, SiO2 aerogels are the most common. This is due, in part, to

the relative simplicity, safety, and reliability of the manufacturing process. Also, many

materials can be incorporated into the aerogel’s structure by involving them in the gelling

process, which helps to tailor the properties of the resulting sample. For instance, a new

approach gives aerogels a much higher compressive modulus and improved tensile

strength while still maintaining the incredible low-densities of the aerogels [17].

6

Aerogels are used in such varied fields as: electronics, life-sciences, space

exploration, green technology, and many others. They were used in JPL’s recent Pheonix

Lander mission to Mars as thermal insulation on the Lander. Aerogels are predicted to be

the next generation interlayer dielectric for integrated circuits [2, 18], paving the way for

faster computing. They are being used in high-efficiency windows [22], and to clean up

oil spills [21]. New technologies and ideas such as these provide efficient ways to

manage energy and ensure the future heath of the environment. It is clear that aerogels

are creating breakthroughs for clean, renewable energy and energy conservation, as well

as environmental-protection technologies.

1.2: Motivation The main motivation for this project is that it will expand students’ access to a

material whose applications are far from exhausted. The reliable synthesis of SiO2

aerogels in a small liberal arts college setting would give future students at Pomona

College and other colleges the opportunity to produce and experiment on this

extraordinary material. Aerogels created in such a setting could also be donated to other

colleges and high schools for use in introductory materials science classes, such as the

one this author took at Olympia High School. The hands-on nature of such classes allows

students to explore science in both theory and application, and aerogels fit perfectly into

many demonstrations suitable for students of all levels.

Beyond their use in the classroom, aerogels produced at small colleges can give

students the chance to explore a material that is emerging as a cutting-edge material in

green technology. This is a field spawned by the fact that the general warming of our

7

planet due to human-produced greenhouse gasses threatens the future stability of our

environment. At the same time, the future of reliable energy is at risk due to dwindling oil

fields and the ever-increasing demand for petroleum products. Incorporation of aerogel

has proven time and time again to increase the energy efficiency of many machines and

systems, thus helping preserve energy supplies by reducing usage. As well, aerogel can

be used to extract pollutants from water, allowing it to capture the harmful particles

before they enter ecosystems [30]. Scientists have realized that they need to do much

more to develop technologies that lead to clean and renewable energy and the

conservation of that energy. Small liberal arts college students can take part in such

research, using this project as a basis to create SiO2 aerogels.

Another course that research using these aerogels can take is the development of

materials and technologies for space-exploration technology. Aerogels have been

researched for fuel-storage tanks, cryogen-storage tanks, and have been implemented for

such applications as the comet-dust catching structure in NASA’s Stardust spacecraft [3].

In the spring of 2008 the author of this paper worked on synthesizing silica-based

aerogels. Two recipes were found online at the Lawrence Berkely National Laboratory

website. Both recipes incorporated the most recent technique in aerogel manufacturing,

replacing TMOS with TEOS as the precondensed silica base. While both recipes

succeeded in creating gels, no gel was successfully evacuated of all liquid and, therefore,

no aerogel sample was produced. However, the necessary equipment and chemicals were

either already owned or were inexpensive to order. In other words, all of the materials

required for the process were in place, assuming corrections to the synthesizing process

could be instituted to allow the production of actual aerogel samples.

8

The most “successful” run from 2008 resulted in a 2.2 cm long piece with a

density of about 0.2 grams per cubic centimeter, which is too dense to be considered an

aerogel. The sample had a rubbery, moist feel, and the smell of ethanol was apparent.

This suggested that the problem lay within the liquid CO2 introduction phase of the

critical point drying step. One possible solution was thought to be to allow the sample to

sit at 0 degrees Celsius for longer than the 2 minutes asked for by the Tousimis critical

point dryer manual [27]. The additional time was hypothesized to allow the ethanol

involved in the gel sample to be replaced by the liquid CO2 present in the rest of the

chamber. After contacting the Tousimis Company, this hypothesis was partially

confirmed. They suggested letting the sample bath two separate times for 15-18 hours,

instead of 2 minutes. The 15-18 hour baths were left at ambient temperature, instead of

being cooled to 0 degrees Celsius as they had been. After each bathing period, two steps

of the process were repeated, as is described in section 4.5.

Fig.1.2 Above: A shrunken, low-density silica sample from Physics 174

Another problem encountered in 2008 was that the molds were not waterproof

and the gels would not form in them, so instead the gels were allowed to form in beakers.

The resulting gels were too large for the critical point drying chamber and breaking them

9

apart before synthesizing damaged their surfaces and possibly their internal structure.

Because of this, the development of better molds was successfully taken on for this

project. Hollow Teflon cylinders with screw-on caps were developed (see Appendix I).

Due to their inert and non-stick nature, the molds allowed for continuous, smooth-

surfaced gels to be formed for this project. This ensured that the surface was unmarred

and that the interchange of liquids at the surface was equal across the entire sample.

1.3: Methods There are three main methods for aerogel fabrication, representing three unique

solutions to the surface tension problems involved in the process. The two methods that

will not be used are rapid supercritical extraction and ambient pressure drying. Though

rapid supercritical extraction is a much faster method than the one proposed here, it also

involves specialized pressure molds and equipment that the Pomona College Physics

Department does not own. The ambient pressure drying method is an elegant solution,

which involves chemically altering the surface of the gels to relieve tension and allow for

drying at ambient pressures, but involves chemicals not currently on hand as well as

difficult-to-reproduce purification processes. The third method, which shall be used here,

is the slower supercritical drying process. This method requires either an autoclave or a

critical point dryer. Professor Tanenbaum’s research lab owns a Tousimis Company

Samdri-PVT-3D critical point dryer.

The three goals for this project were as follows: First, synthesize reproducible,

stable aerogels. This required several important steps: The critical-point dryer’s CO2 inlet

valve was broken and had to be fixed before experiments could take place. The

department ordered a valve replacement from the Tousimis Company and Glenn Flohr

10

executed the repairs. Another step was to research whether or not the planned recipe

would produce hydrophilic or hydrophobic aerogels, and what steps would need to be

taken to make hydrophilic samples stable. It was found that the aerogels made from the

recipe should be innately hydrophobic upon drying and would hold up to air moisture

indefinitely [19]. The second goal for this project was to characterize the resulting aerogel

samples. This was carried out for several physical and optical properties of the samples using

methods described in Chapter 5. The third goal was to consolidate the research performed

herein into a laboratory manual for future Pomona College and other small liberal arts college

students to follow. To view the student manual for synthesizing simple SiO2 aerogels, please

see Appendix III.

11

Chapter 2: Theory 2.1: Chemical Theory

Aerogels begin life as a silica (SiO2) gel. Though aerogels have also been created

with alumina, chromia, tin oxide, carbon and other materials, silica versions are the

easiest and most reliable to produce. The silica gels are about 99% liquid by weight, but

contain matrices that give the liquid surface tension. In a silica gel these molecular

networks are made up of silica dendrites.

The first step in making aerogels is to create what is called a “wet gel.” Currently,

the best way to create the wet gels is to begin with a silicon-alkolyde precursor, such as

TEOS, which was used for this project. The chemical makeup of TEOS is

Si(OCH2CH3)4, which, when added to water, achieves the chemical reaction.

Si(OCH2CH3)4 (liq.) + 2(H2O) (liq.) → SiO2 (solid) + 4(HOCH2CH3) (liq.) eq.2.1 [9]

The amount of water indicated in eq.2.1 is only enough to exactly balance the

reaction. In practice, anywhere from 4 to 30 times more water should be used to increase

the strength of the gels [9]. This chemical reaction is generally accomplished in ethanol.

Acid or base catalysts are added to this reaction in order to decrease gelation time

from several days to around an hour. Base catalysts were used in this project due to their

tendency to create clearer aerogels. Base-catalyzed aerogels also retain more of their

volume throughout processing than do their acid-catalyzed counterparts.

The current precursors are largely alkoxides M(OR)n which are compounds

consisting of a metal M and an alkoxide group OR [19]. With silicon precursors the R

designating the alkyl group is often a methyl CH3 group or an ethyl C2H5 group. The

reason silica-based aerogels have been studied much more than any other type is because

12

silicon atoms carry a reduced partial positive charge δ+. In Si(OEt)4 δ+ is about 0.32 [19].

This reduced partial positive charge makes the gelation kinetics of the silicon precursor

extremely slow. Because of this, the speed of the hydrolysis rate, compared to the

condensation rate, can be controlled through catalysts. Controlling these kinetics allows

silica aerogels to be adapted at this stage for their future application [19].

This project made use of Silbond H-5, a pre-polymerized TEOS silica source. It is

created by adding heat and an acid catalyst to a solution containing TEOS, ethanol and a

sub-stoichiametric amount of water [9]. The resulting fluid is has higher molecular

weight silicon alkoxy-oxides. Because of this, Silbond H-5 only requires processing with

a basic catalyst in order to achieve gelation. The resulting gel is clearly a combination of

a silicon dioxide solid and the liquid HOCH2CH3, as well as left-over ethanol, water, and

Silbond H-5 that did not undergo the reaction. It is the removal of all leftover liquids that

turns the wet gel into an aerogel and gives the material its unique physical properties.

2.2: Supercritical Fluid Theory If a wet gel were to dry at normal room temperature and pressure, the liquid

within the gel would evaporate into a gas and the lattice-structure that gives the gel its

shape would shrink and crumble due to surface tension. The deforming comes from the

capillary pressures, which arise from the diminishing meniscus of the liquid. When these

pressures build up in the lattice’s pores they destroy the lattice. This problem is solved by

introducing the sample into a temperature and pressure controlled environment, such as

an autoclave or a critical point dryer. The pressure in the sample’s environment is raised,

followed by the temperature, until the liquid turns into a supercritical fluid. The point at

which this happens is called the critical point - the point at which a gas cannot be turned

13

into a liquid simply by increasing pressure. At the critical point the fluid will lose all

surface tension properties associated with a liquid and become completely miscible. With

small changes in pressure, the supercritical fluid experiences large changes in density.

Because the liquid-gas interface is eliminated, and therefore no surface tension is present,

the sample can be dried without deformation.

In this project, carbon dioxide was introduced to the sample in liquid form,

replacing the pure ethanol which had been in the gel. The liquid CO2 was then brought

past its critical point of 304.1 K and 72.8 atm. The pressure of the supercritical fluid was

then lowered, while maintaining higher than critical temperatures in order to take

advantage of its properties of low viscosity and high diffusivity. In this manner the

supercritical fluid becomes more gas-like and is vented off, leaving the silicon-dioxide

lattice structure intact.

2.3: Physical Theory The solid nanoscale silica dendrite networks of silica-based aerogels give them

their unique properties. The ionic fraction of the polar covalent bonds for several

different metal oxides yields the following results.

fionic = 1 – exp(-0.25(XM – XO)2) eq.2.2 [19]

XO and XM represent the Allred-Rochow electronegativities of O and M. For

Al2O3, TiO2, ZrO2, and Na2O, fionic gives 0.64, 0.70, 0.71, and 0.78. SiO2 on the other

hand has an fionic of 0.54, allowing the Si-O-Si angle value to range widely, which gives

the resulting structure a “random network” [19]. The other four oxides given have higher

ionic fractions, leading to a lower range of bond angle values. This means that the

random bonding only occurs on a more macro scale with bigger, denser, colloidal

14

particles. When this happens the resulting gel is particulated instead of forming a random

network of dendrites. This means that achieving gelation with higher fionic valued oxides

is difficult and even impossible with a value as high as the 0.78 of Na2O [19].

The silica dendrite network which makes up the aerogels creates an extremely

strong material for its relative weight. The same network also makes aerogels brittle,

causing the material to shatter under sharp pressures. Aerogels are amazing thermal

insulators. The reason for this property is threefold. The air inside of the silica lattice

cannot pass through the nanoscale pores, so convection cannot occur. Silica has a very

low heat transfer coefficient and because of this the heat conductivity through the

physical solid matrix of the aerogel follows the formula for solid conductivity λs (eq.2.3).

λs ~ ργ eq.2.3 [19]

Where ρ is the density of the aerogel, and γ has a value between 1.2 and 1.8.

Lastly, small amounts of carbon can be added to the aerogel, absorbing infrared radiation

and all but halting radiative heat transfer. Because of this, aerogels have thermal

conductivities of 0.03-0.004 W/(mK), up to 6.25 times lower than air. At best, aerogels

have thermal conductivities of up to 10 times lower than wood or polyurethane foam

insulation [19]. Due to this fact, windows no longer need to be the weak point in energy

efficiency for buildings if the windows incorporate aerogels. Out of any currently known

materials, aerogels hold the record for the lowest thermal conductivity, refractive index,

sound velocity, dielectic constant, and bulk density.

The gels produced in this project used Silbond H-5 as the precondensed silica

precursor. Silbond H-5 is derived from TEOS whose chemical formula is Si(OCH2CH3)4.

Due to the non-polar chemical group CH3-Si bond, that accompanies the normal CH2

15

ethyl group, the resulting aerogel is hydrophobic [19]. This allows the aerogel to be

exposed to water and not have its structure collapse due to the pressures involved.

Because of this trait, intrinsically hydrophobic aerogels are more appropriate for a small

liberal arts college setting where obtaining exotic chemicals to treat the surface, or

providing moisture-free environments in which to conduct experiments can be

prohibitively expensive and difficult to create.

16

Chapter 3: Previous Research The area of aerogel research expands enormously each year and profiling a

complete history of previous research done in the area is impossible. Instead, we

document here breakthrough research that opened the door to entirely new applications

based on different properties of the aerogels. Highlighted is research done on a small

scale, as well as research done with silica-based aerogels. This will have the most uses

for this project as the research can be mimicked or examined further in a small liberal arts

college setting.

For further inspection of application and synthesis techniques, one can turn to

Pierre and Pajonk, who have developed new catalysts by using aerogels, as well as

extensively cataloguing previous methods and applications of aerogels. In their 2002

paper, Chemistry of Aerogels and Their Application, they put emphasis on the fact that

aerogels should be considered a different class of material than have ever been seen

before. They provide an in-depth analysis on the chemical processes behind the creation

of different types of aerogels. The paper provides both the history and chemical process

behind each step in aerogel fabrication, as well as the uses of the end-product. Because of

this, Pierre and Pajonk to date have provided the most exhaustive background for any

small liberal arts college aerogel researcher. Most importantly, it maps out the creation of

carbon aerogels, an important material due to its electrical properties. This paper also

details how aerogels are being used in catalysts and the life-sciences, which will provide

a new angle for further research in a small liberal arts college setting [19].

The first method for “removing water from a jelly without an accompanying large

shrinkage” [13] was devised by Steven S. Kistler in 1931 on a very small scale with

17

relatively inexpensive materials. He followed up with studies on Thermal Conductivity of

silica Aerogels in 1938, and Sorption and Surface Area in Silica Aerogel in 1943 [13,

14]. Though modern-day aerogels are produced through different methods and materials,

and are manufactured to exhibit much more specialized properties, it was obvious to

scientists like Kistler from the very beginning that aerogels exhibited extraordinary traits

including low density and low thermal conductivity.

After half a century of relatively little study, aerogels came back into the research

spotlight in the early 1980’s as a material suitable for employment as a Cerenkov

radiator. Carlson et al utilized the refractive index (n = 1.01 to 1.02) and the transparent

nature of silica aerogels to incorporate them in a Cerenkov detector design. Aerogels

represented a breakthrough for Cerenkov detector designs because the detectors no longer

had to involve liquefied hydrogen or helium to provide the correct indices of refraction

[1].

Many researchers look to utilize one or two of aerogels’ record-breaking

properties. In 1992, Tillotson and Hrubesh examined the possibility of maximizing the

low density and transparency of aerogels by using a two-step process. This two step

process is performed with tetramethylorthosilicate as the precondensed silica precursor in

their paper Transparent Ultralow-Density Silica Aerogels Prepared by a Two-Step Sol-

Gel Process. However, this method can be done just as well with tetraethylorthosilicate

(TEOS) and is one of the most widely-used processes to develop low-density highly-

transparent aerogels 17 years later. The authors also attempted to characterize aerogel

samples with a scanning electron microscope. They found that while electron-charging on

the surface of the aerogels makes high-magnification, high-resolution images impossible,

18

lower magnification images yield high enough resolution to be useful in showing the

various microstructures of aerogels synthesized through different processes [26].

In 2001, Reynolds et al utilized the absorption properties of aerogels in order to

create a cheap, quick, and environmentally safe method of cleaning up crude oil spills in

salt water. Hydrophobic aerogels functionalized with CF3 groups were used to absorb oil

from an oil-seawater solution. After absorption, both the water and the oil were purified

and could be reused. This experiment took place on a scale that is easily reproducible,

though the exotic nature of the aerogels makes them hard to fabricate. This research also

has the benefit of using aerogel powders over solid monoliths, a fact that lowers the cost

of materials significantly. The author’s publication of Hydrophobic Aerogels for Oil-Spill

Cleanup - Intrinsic Absorbing Properties led to a patent for their environment-saving

aerogel technology [21].

The theory for further green technology based on aerogels was developed in 2001

by Krainov and Smirnov. The scientists hypothesized that absorbing deuterium in SiO2

aerogels would yield an improved target for inertial confinement laser fusion. Laser

Induced Fusion in Aerogel details the theoretical arguments behind their suggestion and

shows that the fractal nature of the aerogel skeleton should lead to a “Coulomb explosion

of the fibers” [15], increasing the power generated through fusion by a factor of ten. The

theory developed in this paper will soon be put to test at the just-finished National

Ignition Facility at the Lawrence Livermore National Laboratory in Livermore,

California. The aerogels used for this theory were simple SiO2 aerogels, unveiling yet

another way in which aerogels might help advance green technology, this time by

providing a theoretical solution for clean, renewable energy [15].

19

The dielectric properties of aerogels led to them being used in commercial

capacitors. By 2002, the PowerStorR aerogel capacitors, which use carbon aerogels as the

electrodes, were up to 5000 times smaller for a given capacitance when compared to

normal electrolytic capacitors. Juzkow diagramed the design process of the aerogel

capacitors. Following the methods presented, it should be relatively easy to create and

test aerogel capacitors on a case-by-case basis. The only requirement would be to make

or obtain carbon-based aerogel samples, as they are best suited for use as electrodes [12].

The dielectric properties of aerogels also give them a potential future in integrated

circuits, as long as several key hurdles can be overcome. In 2005, Cho et al examined a

process to seal the pores at the surface of the aerogels in order to make them useful as

interlayer dielectrics without compromising the other important properties of the

aerogels. The authors found that if the pores were not sealed before CVD processes

involved in making integrated circuits were carried out, then the entire aerogel interlayer

dielectric became contaminated and useless. The solution presented was to seal the pores

by exposing them to CHF3 plasma. The model created predicted that a one-minute

treatment in a 90 mTorr CHF3 plasma would suffice to seal the pores. If hydrophilic gels

were to be created in a small liberal arts college setting, then this method of pore sealing

could allow further experimentation on the gels, without having to provide a moisture-

free environment. The only requirement would be the ability to create a 90 mTorr CHF3

plasma that a sample can be exposed to for one minute [2].

In their 2007 paper Effective Preparation of Crack-Free Silica Aerogels via

Ambient Drying, Hwang et al detail experimental techniques and designs used to make

solid silica aerogel preparation safer and less costly - the two main hurdles that must be

20

overcome before the potential of aerogels becomes fully recognized. The research skirts

the normal technique of using expensive chemicals, such as alkoxides, and the expensive

and somewhat dangerous method of supercritical drying. With the idea to create

inexpensive aerogels for the scientific community, the authors used waterglass as the

base material, a low-cost replacement for tetraethylorthosilicate (TEOS) and other

chemicals. They also developed a technique to dry the aerogels at ambient pressures, and

overcome surface tension problems through surface modification. As long as the

inexpensive, but intricate ion-exchange system built by the authors to purify the silica

solution can be reproduced, this process should be easy to mimic [10].

In 2008, Meador et al explored a new approach to synthesizing aerogels that

resulted in samples with a much higher compressive modulus and tensile strength than

previously achieved. In order to maintain the desirable low-densities of aerogels, carbon

nanofibers were included in the gel-process, allowing them to become involved with the

resulting silica aerogel dendrite network. Though the chemical reaction involved in this

paper is not identical to this author’s process, the process of incorporating carbon

nanofibers can be appropriated to provide a platform for further research with the Silbond

H-5 process. The carbon nanofibers are processed through sonication into the liquid silica

solution. The authors also detailed how to test the tensile strength of fabricated aerogels

[17].

One aerogel technology that has been developed and awaits only a cheaper mode

of production is the integration of aerogels into modern high-efficiency windows.

Aerogel windows make use of the thermal-conductivity and transparent properties of the

material. This technology was implemented in a standard Danish house in 2008 and was

21

found to reduce annual energy needs by 19% [22]. When implemented in “low-energy”

Danish houses, the aerogel windows reduced annual energy needs by 34% [22].

Other ongoing research includes producing aerogels in a weightless environment,

because they are much more transparent than their “blue smoke” brethren and would

allow for insulating windows with more favorable optical properties. Aerogels are also a

common material in spacecraft insulation and were famously used as the sample-

collection material on NASA’s Stardust spacecraft, which collected specimens from the

tail of the comet Wild 2. The comet dust created a trail as it broke through the silica

lattice, before stopping at some point in the aerogel structure. Due to the strength, the

varying density, and the transparent quality of the aerogel structure, scientists were able

to pinpoint the position of the comet dust and study it when Stardust returned [11].

22

Chapter 4: Experimental Setup and Process 4.1: Materials

The Silbond H-5 Company donated 500 mL of the Silbond H-5 precondensed

silica solution to the author. Though it is not sold in such small quantities, Silbond H-5 is

similar to TEOS, which goes for about $50 per liter. The physics department had enough

pure ethanol on hand for both the recipe and the following “baths”. The water was

obtained from the Millipore Direct Q 3 UV water filter in the Physics department. The

30% aqueous ammonia solution was acquired from the Chemistry Department. One 100-

1000 uL pipette was borrowed from Professor Kwok. The second 100-1000 uL pipette

and all other glassware was bought or borrowed from the Chemistry Department. The

Teflon was purchased by Glenn Flohr. The screws in the molds were fitted by Glenn

Flohr. The Tousimis PVT-3D critical point dryer is owned by Professor Tanenbaum’s

research lab.

4.2: The Gel Stage The first step in synthesizing aerogels was to create a gel. This thesis made use of

the “two-step acid-base catalyzed silica aerogel” recipe from the Lawerence-Berkely

National Laboratory website [9]. The recipe was prepared underneath a fume hood for

safety. The steps were as follows.

1) In a 500 mL beaker 25 mL of Silbond H-5 was mixed with 25 mL of ethanol. 2) The catalyst solution was mixed in a 100 mL graduated cylinder; its

components were 17.5 mL of ethanol, 37.5 mL of water, and 0.175 mL of 30%

23

aqueous ammonia. The 0.175 mL of 30% aqueous ammonia required the digital

pipette due to the small amount of liquid.

3) The catalyst solution was poured slowly into the Silbond H-5 solution stirring

the entire time. The pouring took a minimum of 45 seconds. Any quicker and a

foam developed on top of the mixture, compromising future gels.

4) The resulting mixture was then poured into the Teflon molds. 4.3: The Mold Stage

The LBNL website claimed that this recipe would take 30-90 minutes to gel,

assuming no temperature extremes. The gels created in this paper were gelled by the 90

minute mark. After the gels were set, the following solution was prepared as suggested by

the website on William Wood’s aerogels [24].

17.5 mL ethanol. 37.5 mL water.

The molds holding the samples were then set in 50 mL beakers and completely

covered with the new solution. This kept the samples from being exposed to air which

would dry them out and destroy the silicon lattice structure [24]. The 50 mL beakers were

then covered with tin foil in order to keep the evaporation rate of the covered solution

low. The gels were allowed to set in the molds for over 48 hours in order to have the

lattice structure fully develop.

4.4 The Bath Stage

After the gels were fully set, the next step was to remove them from the molds

and set them in ethanol baths. A 50 mL beaker containing 25 mL of pure ethanol was

prepared for each gel. The molds were then extracted from their previous beakers, the

24

bottoms unscrewed and the gels pushed gently into their individual baths. The baths were

drained and replenished with an equal amount of ethanol every 8 to 24 hours. Each

sample was introduced to a total of at least 5 pure ethanol baths in order to ensure that the

liquid held within the gel matrix was pure ethanol with no water. Because the volume of

pure ethanol in each bath was 1.5 times the volume of the gel (the gel recipe is 35.7%

water by volume), the final sample should not have contained more than .4% water by

volume. At this point the gels were ready for the critical point dryer step. The gels

appeared absolutely clear in the beakers and had only several small bubbles in them

acquired in the mixing stage. While able to be moved by hand, the more contact with

pressure and air, the higher probability that the gel would weaken, develop cracks, or

even break. The best method was to leave the gels in one beaker, and if they had to be

moved, to slide them gently from beaker to beaker. Even when care was taken in every

step to minimize pressures and stresses endured by the gel, about half the samples still

broke, fell apart, or did not form a coherent gel matrix.

4.5: The Drying Stage The Tousimis Samdri PVT-3D critical point dryer was used in the drying stage of

this process. The critical point dryer chamber contained two openings, the liquid CO2

entry line and the exhaust line. Because the gels could potentially run into these openings

and block up the lines, filter paper was cut and folded into a cylindrical pipe with one

bottom. Because it did not interact with the CO2 or ethanol, the filter paper was able to

effectively block the two openings from gel particulate, while remaining invisible to the

other chemicals in the experiment.

25

After the filter paper was in place, lining the drying chamber, ethanol was poured

into the chamber until it was half full. Then the gel sample was extracted gently from its

bath and placed into the drying chamber. The gel was then processed according to the

Tousimis Advanced Manual Critical Point Dryer Samdri-PVT-3D user manual [27].

However, there was one important difference between the instructions and the procedures

used in this project. As per Dr. A.J. Tousimis’ suggestion, after step 7 was accomplished,

the sample was allowed to sit for 15-18 hours without moving on to step 8. After that

time period, steps 6 and 7 were repeated and the gel was allowed to sit in the same

manner as before for 15-18 hours. During these wait times, the chamber did not need to

be kept below ambient temperature. After the second waiting period, steps 6 and 7 were

repeated for a third time and then the procedure continued as normal with step 8 until the

end.

When repeating steps 6 for the second and third time, it is imperative that viewing

chamber be watched to guarantee that the liquid CO2 level does not drop below the

sample. This misstep is easy to do because the liquid CO2 is cooling quickly from above

ambient temperature to around 0 degrees Celsius. While cooling, the pressure drops

significantly, and if it drops below 850 psi, a gas pocket develops in the chamber. If the

liquid drops below the sample, allowing the gas pocket to come in contact with the

sample, then the sample quickly develops cracks and begins to turn opaque. This change

in appearance represents the deforming of the solid matrix within the gel, and will result

in shrunken, denser, and highly-weakened samples. The critical point dryer, including the

viewing chamber, can be seen in Appendix I.

26

Each time, when performing step 6 from the Tousimis critical point dryer manual,

the beaker connected to the exhaust line should be monitored. When opaque white solids

enter the beaker, looking like snow, then the critical point dryer chamber has been

evacuated of ethanol and only liquid CO2 is left, except for any ethanol involved in the

gel matrix. This should only take about 10 minutes of purging and replacing the liquid in

the chamber. An image of the exhaust beaker can be found in Appendix I.

At the point when the CO2 was extracted for the final time, the gel sample became

an aerogel, with only the solid silicon matrix of the gel left over. Each sample was

removed from the chamber, measured, weighed, and placed into a container, ready for

further testing.

27

Chapter 5: Results and Analysis

5.1: Qualitative Results

In 2008 the author attempted to create aerogels, but only succeeded in creating

shrunken low-density silicon dioxide solids that still had ethanol inside them. The best

result was a 2.2 cm long piece with a density of about 0.2 g/cm3, or about a factor of 3

larger than expected from the recipe. When crushed between two fingers, the sample felt

rubbery and moist and the smell of ethanol was evident. Since the sample had been

bathed 8 times in ethanol equal to its volume, the problem was not with water being

involved in the matrix. Instead, the presence of the ethanol pointed to the fact that that the

sample had not bathed long enough during the liquid carbon dioxide replacement stage of

the critical point drying phase. The sample shrank continuously from the time of

extraction and initial measurement until about 24 hours later when it was about half the

size and much more dense.

Fig.5.1 Above: Failed aerogel sample from Physics 174

28

The shrinking of the failed sample led the author to the initial, incorrect

conclusion that the recipe produced hydrophilic aerogels, whose solid matrix would

deform due to moisture in the air. However, successful samples proved this hypothesis

incorrect, showing no shrinking after weeks of contact with the air. The first successful

sample was roughly in the shape of a cylinder 2.40 cm in height and 1.34 cm in radius.

With a weight of .3115 grams, the sample had an overall density of .079 g/cm3, exactly

on par with the .08 g/cm3 that the recipe predicted.

Fig.5.2 Above: First successful aerogel from the improved method The optical properties of the aerogels produced herein included the following:

The sample was generally clear, with some white opaqueness when laid against a white

background. When laid on a dark background, especially a black background, the sample

took on a smoky, blue tinge. Cracks ran along the height of the cylinder derived from one

of the last two liquid CO2 replacement stages.

29

The sample produced in the next run resulted in a failed aerogel. Because the

result turned out qualitatively identical to the failed samples from the Physics 174 project,

it pointed to the fact that the sample failed to exchange all of its pure ethanol for liquid

CO2, even after the two 18-hour baths in the critical point drying phase. The baths did not

suffice because the sample was from the larger of the molds, which was designed to

produce a gel with a volume just smaller than that of the critical point dryer chamber.

This meant that not enough liquid CO2 could ever be present to replace all of the pure

ethanol present in the sample. Because of this result, later gels produced using the

maximum-capacity mold were cut in half.

The second successful batch of aerogels consisted of sections from two gels that

were broken when being extracted from their molds. The point of this run was to see if

broken gels could still produce useful samples. The results were interesting: The viewing

chamber was monitored at the point when the liquid CO2 levels fell below the level of the

top-most samples. The exposed portions of these samples immediately developed an

opaque white appearance and developed deep cracks. When the CO2 level was brought

back up, the samples stopped degenerating. It is clear from Fig.5.3 which of the smaller

samples was exposed to the gas pocket in the critical point dryer. One small sample to the

left of the screw has a single opaque white face where the liquid momentarily dipped

below the sample. The sample at the bottom left-hand corner of the cylindrical aerogel

sample was the topmost sample in the chamber and was fully exposed to the gas pocket

for several seconds, giving it its deeply cracked, opaque, and eroded look.

30

Fig.5.3 Above: Aerogel samples and a screw on the hot plate Due to the amazing capacity of aerogels to be thermal insulators, this property

was the target of several tests carried out in this project. The first was to program a hot

plate to 150 degrees Celsius and place the aerogels samples on top. After allowing the

samples 10 minutes to equilibrate, the aerogels were imaged alongside a metal screw with

a Fluke thermal imaging camera. Even though the screw’s emissivity was not the same as

the aerogels, and thus appeared cooler than it actually was, it still showed higher heat

conduction than the large aerogel sample. While it was difficult to focus well enough on

the small samples to see their heat conduction properties, the large sample demonstrated

this trait extraordinarily well, remaining at a much cooler temperature than the hot plate

and showing almost no thermal gradient within its height.

31

Fig.5.4 Above: the larger aerogel sample alone on the 150 degrees Celsius hot plate

Below: the larger sample accompanied by the smaller samples and a screw on

the 150 degrees Celsius hot plate

Fig.5.5 The screw pictured in fig.5.5 seems to be at only a slightly higher equilibrium

temperature than the aerogel. However, the screw and the hot plate were both made of

reflective material, so that their emissivity was not the same as the aerogel. Because of

this, and the fact that the hot place was programmed to be at 150 degrees Celsius, it is

32

clear that both the screw and the hot plate were actually 10 - 20 degrees Celsius warmer

than they appear in the image.

Fig.5.6

Above: aerogel sample on the makeshift hot plate

In the second test the aerogel was heated up with an oxy-acetylene torch. Direct

exposure to the torch caused the surface of the aerogel to begin deforming. Because of

this, a makeshift hot plate was constructed (Fig.5.6) that would allow the aerogel to be

about 100 degrees Celsius hotter than the commercial hot plate, but would not deform the

aerogel. It is clear from Fig.5.7 and Fig.5.8 that even under temperatures hot enough to

begin bending the steel plate it sat on, the aerogel did allow heat to transfer along its

height.

33

Fig.5.7

Above: Thermal side view of the aerogel on the heated makeshift hot plate

Fig.5.8

Above: Thermal top view of the aerogel on the heated makeshift hot plate

Due to their low densities of 0.08 g/cm3, the silica solids produced in this thesis

classify as aerogels. Below are scanning electron micrographs of an aerogel produced in

this project, and a commercial aerogel owned by the Pomona College physics

department. While charging of the aerogels made anything beyond 50,000 times

magnification impossible, it is easy to see that both solids exhibit nanoscale structures.

34

The lattice and porous nature of this structure is evident, agreeing with the theoretical

ideal that aerogels are almost completely air by volume.

Above Left: Fig.5.9. Thesis sample at 30,000x mag. [23]

Above Right: Fig.5.10. Commercial sample at 20,000x mag. [23] Further magnification does help to show the 3-dimensional nature of the lattice

structure. Below, the crater at the surface of an aerogel produced in this project shows the

lattice structure as a continuous 3-dimensional network as was expected. The individual

dendrites are on the order of several nanometers wide. The right-hand image shows the

same magnification of a commercial sample, which exhibits a homogenous surface of

pores and dendrites.

35

Above Left: Fig.5.11. Thesis sample at 50,000x mag. [23]

Above Right: Fig.5.12. Commercial sample at 50,000x mag. [23]

36

Chapter 6: Conclusions and Future Work 6.1: Conclusions The questions and hurdles involved in devising a method to create simple SiO2

aerogels in a small liberal arts college were met and resolved by this project. Future

students at Pomona College will have access to the attached manual (see Appendix III),

which outlines the steps to produce silica aerogels. With this, they can bypass the

problems encountered by this author. Though the method takes about a week from start to

finish, the time that the student must be in the laboratory is only about 6 hours in total.

While the larger samples produced herein exhibit cracks, many of the smaller ones do

not, and all samples held up during experimentation.

New technologies and ideas such as these provide solutions that will help improve

the future of energy and the environment. Cheap, reliable access to aerogels for

undergraduate students, in particular, can help increase interest and research into clean,

renewable energy and energy conservation, as well as environmental-protection

technologies.

6.2: Future Work

The sustained application of this project lies in the utilization of this method by

small liberal arts college students to create SiO2 aerogels as a springboard for future

research and application.

One application that would be useful to lower-level physics courses, including

high school classes, would be to develop demonstrations or lab experiments with the

aerogel samples. Demonstrations that could potentially fill this niche are varied: A

demonstration of the thermal insulation properties of aerogels could include setting a

37

piece of paper, or a match on top of a piece of aerogel, and then holding a blow-torch or

Bunsen burner below. The idea would be to examine how heat travels through different

materials, with aerogels being the least-conductive of any material presented. Because the

strength-to-weight ratio of aerogels is very high, a demonstration or laboratory could be

developed to show why, and how, aerogels were used to capture comet dust in NASA’s

Stardust mission. This might include weighing the aerogel and then gently placing many

times its own weight on top of it in the form of obviously heavy materials, such as metal,

or a section of wood. One problem with developing laboratories will be the propensity for

the aerogel samples to shatter if handled incorrectly. Because of this limitation, CF3-

functionalized aerogels could provide a great lab that does not involve moving or

touching the aerogel samples. The CF3-functionalized aerogels would allow the creation

of a lab that would show the water purification capabilities of the samples [21]. Another

lab could show why aerogels are replacing conventional windows in some instances,

even though the aerogels are chemically identical to, but more expensive than, everyday

glass. The lab could provide images of the differences between lattice networks of the

low-density aerogel versus an image of the compact, non-porous glass surface. The

difference in electromagnetic radiation transfer through glass and aerogel would bring the

point home [22].

Another goal of future research should be to execute original experiments.

Because of this, the molds for the samples were made to yield cylindrical samples of

scalable height. If little of the solution is introduced into the mold, then a thin disk will result,

which will aid in the fast production of the aerogels and increase their usefulness as a

platform for further experimentation.

38

One experiment that should be possible would be based off of the research of

Meador et al that incorporated carbon nanofibers into the aerogel structure. Because of

this research, it seems clear that basic SiO2 aerogels should serve as a good platform on

which to grow carbon nanotubes [17]. Meador et al also showed that involving carbon

nanofibers in aerogel production yields aerogels with similar densities, but much higher

tensile strengths and compressive modulii than unaltered samples. This mans that it might

be possible to involve carbon nanotubes in the chemical process of creating silica

aerogels, just as Meador et al included carbon nanofibers, because the two materials are

identical in chemistry. If, for instance, aerogels could be grown around a field of carbon

nanotubes, it might provide a use in electronics as a super-strong, super-small circuit

material. The method of incorporating the carbon nanofibers – sonication into the liquid

silica solution – could be used in the case of carbon nanotubes as well, or the aerogels

could be grown around a smaller aerogel sample with nanotube growth already present

on its surface. A scanning electron microscope can be used to examine the surface of the

aerogels and the nanotubes grown on top of it [26]. An atomic force microscope will

likely not be useful in examining the aerogel surface because the porous structure will

preclude high-quality imaging.

Further methods and recipes to create aerogels may be examined. This might

include using different techniques, but the same chemicals, or, using the same critical

point-drying technique, but using different chemicals, such as hydrogen silsesquioxane

(HSQ) [23]. If other recipes are attempted which produce hydrophilic aerogels, then the

methods of Liu et al address the issue of altering the surface chemistry of the aerogels by

treating them with isopropyl alcohol (IPA) and trimethylchlorosilane (TMCS) [16]. This

39

processing should allow the samples to be exposed to air and moisture indefinitely, if this

fails to alter the surface chemistry, then it might be possible to carry out experiments in

an N2-purged glovebox, or some other moister-free environment [23].

40

7: Acknowledgements

I would like to thank Professor Tanenbaum for his guidance and interest in this

project over the past year and a half. A big thank you to Glenn Flohr who has been an

incredible help, both by fixing the critical point drying machine, and spending his own

time to help purchase and thread the Teflon molds. I wish to thank David Haley for his

help in gathering various minor materials for the project, and Professor Kwok for

supplying a digital pipette. I would like to thank Christina Frausto of the Chemistry

Department for her help in acquiring the necessary glassware and various measuring

equipment for this project. I would like to thank Professor Taylor of the Chemistry

Department for providing suggestions and solutions to several of the hurdles initially

faced in this project. I also wish to thank Dr. A.J. Tousimis of the Tousimis Company for

his willingness to share aerogel-synthesis procedures developed by him and his staff in

his lab.

41

8: Bibliography

[1] Carlson, P.J.; Johansson, K. E.; Norrby, J.. 1980. Development of Silica Aerogel

Cerenkov Detectors. IEEE Transactions on Nuclear Science Vol. NS-27, (1):

96-100.

[2] Cho, Woojin; Saxena, Ravi; Rodriguez, Oscar; Ojha, Manas; Achanta, Ravi;

Plawsky, Joel L.; Gill, William N. 2005. Polymer Penetration and Pore Sealing in

Nanoporous Silica by CHF3 Plasma Exposure. Journal of The Electrochemical

Society 152 (6): F61-F65.

[3] Fesmire, J.E.; and Sass, J.P.. 2008. Aerogel Insulation Application for Liquid

Hydrogen Launch Vehicle Tanks. Cryogenics 48: 223-231.

[4] Fricke, J; and Tillotson, T.. 1997. Aerogels: Production, Characterization, and

Applications. Thin Solid Films (297): 212-223.

[5] Gao, Siliang; Wang, Yujun; Wang, Tao; Luo, Guangsheng; Dai, Youyuan. 2009.

Immobilization of Lipase on Methyl-Modified Silica Aerogels by Physical

Adsorption. Bioresource Technology 100: 996-999.

[6] Gauthier, Ben M.; Bakrania, Smitesh D.; Anderson, Ann M.; Carrol, Mary K. 2004.

A fast supercritical extraction technique for aerogel fabrication. Journal of Non-

Crystalline Solids 350: 238-243.

[7] Hostler, S. R.; Abramson, A. R.; Gawryla, M. D.; Bandi, S. A.; Schiraldi, D. A..

2008. Thermal Conductivity of a Clay-Based Aerogel. International Journal of

Heat and Mass Transfer. doi:10.1016/j.ijheatmasstransfer.2008.07.002.

[8] Hrubesh, Lawrence W. 1998. Aerogel Applications. Journal of Non-Crystalline

Solids 225 (1): 335-342.

42

[9] Hunt, Arlon; Ayers, Michael. Silica Aerogels. Microstructured Materials Group –

Lawrence Berkely National Laboratory. http://eetd.lbl.gov/ecs/aerogels/sa-

home.html. 2009.

[10] Hwang, Sung-Woo; Jung, Hae-Hyun; Hyun, Sang-Hoon; Ahn, Young-Soo. 2007.

Effective Preparation of Crack-Free Silica Aerogels via Ambient Drying. J Sol-

Gel Sci Techn 41: 139-146.

[11] Jones, Steven M. 2006. Aerogel Space Exploration Applications. J Sol-Gel Sci

Technology 40: 351-357.

[12] Juzkow, Marc. 2002. Aerogel Capacitors Support Pulse, Hold-Up, and Main Power

Applications. Power Electronics Technology 28 (2): 58-60.

[13] Kistler, S. S. 1933. Coherent Expanded Aerogels. Journal of the American

Chemical Society 65 (10): 1909-1919.

[14] Kistler, S.S. 1938. Thermal Conductivity of Silica Aerogels. Ind. Eng. Chem. 30 (9):

1082-1086.

[15] Krainov, V.P. and Smirnov, M.B. 2001. Laser Induced Fusion in Aerogel. Laser

Physics 12 (4): 781-785.

[16] Liu, Ming-Long; Yang, De-An; Qu, Yuan-Fang. 2008. Preparation of Super

Hydrophobic Silica Aerogel and Study on its Fractal Structure. Journal of Non-

Crystalline Solids 354: 4927-4931.

[17] Meador, Mary Ann B.; Vivod, Stephanie L.; McCorkle, Linda; Quade, Derek;

Sullivan, Roy M.; Ghosn, Louis J.; Clark, Nicholas; Capadona, Lynn A. 2008.

Reinforcing Polymer Cross-Linked Aerogels with Carbon Nanofibers. Journal of

43

Materials Chemistry 18: 1843-1852.

[18] Park, Sung-Woo; Jung, Sang-Bae; Yang, Jun-Kyu; Park, Hyung-Ho; Kim, Hae-

Cheon. 2002. Ambient pressure dried SiO2 aerogel film on GaAs for application

to interlayer dielectrics. Thin Solid Films 420-421: 461-464.

[19] Pierre, Alain C. and Pajonk, Ge´rard M. 2002. Chemistry of Aerogels and Their

Applications. Chem. Rev. 102: 4243-4265.

[20] Pietron, Jeremy J.; Stux, Arnold M.; Compton, Ratonya, S.; Rolison, Debra R..

2007. Dye-Sensitized Titania Aerogels as Photovoltaic Electrodes for

Electrochemical Solar Cells. Solar Energy Materials & Solar Cells 91: 1066-

1074.

[21] Reynolds, John G.; Coronado, Paul R.; Hrubesh, Lawrence W. 2001. Hydrophobic

Aerogels for Oil-Spill Cleanup - Intrinsic Absorbing Properties. Energy Sources

23: 831-843.

[22] Schultz, J.M. and Jensen, K.I. 2008. Evacuated aerogel glazings. Vacuum 82: 723-

729.

[23] Tanenbaum, David. Discussions with author. 2008-2009.

[24] Ten-Year-Old Child Produces Homemade Aerogels. Adzoe

http://adzoe.8m.com/Aerogelsa.htm. 2008.

[25] Timothy, B. Roth; Anderson, Ann M.; Carrol, Mary K. 2008. Analysis of a rapid

supercritical extraction aerogel fabrication process. Prediction of thermodynamic

conditions during processing. Journal of Non-Crystalline Solids 354 (31): 3685-

3693.

[26] Tillotson, T.M. and Hrubesh, L.W. 1992. Transparent ultralow-density silica

44

aerogels prepared by a two-step sol-gel process. Journal of Non-Crystalline Solids

145 (1-3): 44-50.

[27] Tousimis, A. J; Staff. 2005. Tousimis Advanced Manual Critical Point Dryer

Samdri-PVT-3D. 15-17. Tousimis Inc.

[28] Wikipedia. Thermal Conductivity.

http://en.wikipedia.org/wiki/Thermal_conductivity. 2009

[29] Wood, Alan. CO2 Phase Diagram.

http://www.teamonslaught.fsnet.co.uk/co2_info.htm. 12/11/2008.

[30] Xu, Pei; Drewes, Jorg; Heil, Dean; Wang, Gary. 2008. Treatment of Brackish

Produced Water Using Carbon Aerogel-Based Capacitive Deionization

Technology. Water Research 42: 2605-2617.

[31] Yao, Yunjin; Zhang, Suping; Yan, Yongjie. 2008. CVD Synthesis and Purification

of Multi-walled Carbon Nanotubes. 2008 2nd IEEE International Nanoelectronics

Conference (INEC 2008): 562-565.

45

9: Appendix I: Images of Equipment

Above: Maximum-capacity mold and best-sample-producing mold and screwdriver.

Above: Filter paper cut to fold up into a cylindrical cup to hold the gel sample inside the

critical point dryer chamber.

46

Above: Samdri-PVT-3D critical point dryer and liquid CO2 supply canister setup. The

viewing chamber is the central silver disk on the lower half of the dryer.

Above: The exhaust beaker which catches liquid ethanol and CO2 while allowing gasses

to pass through the left tube to the fume hood.

47

Above: The fluke thermal imaging camera.

48

10: Appendix II: More Sample Images

Above: Two areas of thesis sample at 50,000x mag. [23].

Left: Bird’s-eye view of thesis sample. Right: Side view of thesis sample and screw on

hot plate under white light.

49

Left: an image of the bare hot plate. Right: the larger sample alone on the hot plate.

Above: A bird’s eye view of the aerogel on the hot plate

Above: a side view of the larger sample with smaller samples and the screw.

50

Above: A torch in direct contact with the aerogel sample, which caused deformation.

Above: Top-down view of the aerogel on the makeshift hot plate.

51

11: Appendix III

Student Manual for Synthesizing SiO2 Aerogels

In 1931, Steven Kistler made a bet with a colleague that he could prove a wet gel contained a solid matrix the same size and shape of the gel. To do this he began with a gel and extracted the liquid, leaving a low-density solid behind. Using an autoclave to drive the liquid past its critical point he was able to conquer the obstacle of surface tension which would otherwise rip apart the internal solid structure of the gel. His successful wager produced the first silica-based aerogel. Silica aerogels begin life as a silica (SiO2) gel and are about 99% liquid by weight, but contain a network of silica dendrites that give the liquid surface tension. The first step in making aerogels is to create what is called a “wet gel.” Currently, the best way to create the wet gels is to begin with a silicon-alkolyde precursor, such as TEOS, which was used for this project. The chemical makeup of TEOS is Si(OCH2CH3)4 and when added to water, achieves the chemical reaction.

Si(OCH2CH3)4 (liq.) + 2(H2O) (liq.) → SiO2 (solid) + 4(HOCH2CH3) (liq.) eq.1 [9] The amount of water indicated in eq.1 is only enough to exactly balance the

reaction and in practice anywhere from 4 to 30 times more water should be used to increase the strength of the gels. This chemical reaction is generally accomplished in ethanol. The reason silicon-based aerogels have been studied much more than any other type is because silicon atoms carry a reduced partial positive charge δ+. In Si(OEt)4 δ+ is about 0.32 [19]. This reduced partial positive charge makes the gelation kinetics of the silicon precursor extremely slow. Because of this, the speed of the hydrolysis rate compared to the condensation rate can be controlled through catalysts. Controlling these kinetics allows silica aerogels to be adapted at this stage for their future application [19]. This project makes use of Silbond H-5, a pre-polymerized TEOS silica source which only requires processing with a basic catalyst in order to achieve a gel. It is the removal of all left-over liquids that turn the wet gel into an aerogel and give the material its unique physical properties. Supercritical Fluid Theory If a wet gel were to dry at normal room temperature and pressure, then the liquid within the gel would evaporate into a gas and the lattice-structure that gives the gel its shape would shrink and crumble. The deforming comes from the capillary pressures arising from the diminishing meniscus of liquid. When these pressures build up in the lattice’s pores, they destroy the lattice. This problem is solved by introducing the sample into a temperature and pressure controlled environment, such as an autoclave or a critical point dryer. The pressure in the sample’s environment is raised, followed by the temperature, until the liquid turns into a supercritical fluid. The point at which this happens is called the critical point. This is the point at which a gas cannot be turned into a liquid simply by increasing pressure. At the critical point the fluid will lose all surface tension properties associated with a liquid and become completely miscible. With small changes in pressure, the supercritical fluid experiences large changes in density.

52

Above: CO2 Phase Diagram [29]

In this project, carbon dioxide will be introduced to the sample in liquid form, replacing the pure ethanol in the gel. The liquid CO2 will then brought past its critical point of 304.1 K and 72.8 atm. The pressure of the supercritical fluid is then lowered, while maintaining higher than critical temperatures in order to take advantage of its properties of low viscosity and high diffusivity. In this manner the supercritical fluid becomes more gas-like and is vented off, leaving the silicon-dioxide lattice structure intact. Physical Theory The solid nanoscale silica dendrite network of silica-based aerogels give them their unique properties. The ionic fraction of the polar covalent bonds for several different metal oxides yields the following results. fionic = 1 – exp(-0.25(XM – XO)2) eq.2 [19] XO and XM represent the Allred-Rochow electronegativities of O and M. For Al2O3, TiO2, ZrO2, and Na2O, fionic gives 0.64, 0.70, 0.71, and 0.78. SiO2 on the other hand has an fionic of 0.54, allowing the Si-O-Si angle value to range widely which give the resulting structure a “random network” [19]. The other four oxides given have higher ionic fractions, leading to a lower range of bond-angle values. This means that the random bonding only occurs on a more macro scale with bigger, more dense colloidal particles. When this happens the resulting gel is particulated instead of forming a random network of dendrites of particles. This means that achieving gelation with higher fionic valued oxides is difficult and even impossible with a value as high as the 0.78 of Na2O [19]. The silica dendrite network, which makes up the aerogels create an extremely strong material for its relative weight, though they can shatter under sharp pressure. Aerogel is also an amazing thermal insulator. The reason for this is threefold: The air inside of the silica lattice cannot pass through the nanoscale pores, so convection cannot occur. Silica has a very low heat transfer coefficient and because of this, the heat

53

conductivity through the physical solid matrix of the aerogel follows the formula for solid conductivity λs. λs ~ ργ eq.3 [19] Where ρ is the density of the aerogel, and γ has a value between 1.2 and 1.8. Small amounts of carbon can be added to the aerogel, absorbing infrared radiation and all but halting radiative heat transfer. Because of this, aerogels have thermal conductivities of 0.03-0.004 W/(mK), up to 6.25 times lower than air. At best, aerogels have thermal conductivities of up to 10 times lower than wood or polyurethane foam insulation [21]. Due to this fact, windows no longer need to be the week point in energy efficiency for buildings if the windows incorporate aerogels. Out of any currently known materials, aerogels hold the record for the lowest thermal conductivity, refractive index, sound velocity, dielectic constant, and bulk density. The Gel Stage To first step in creating aerogels is to produce a gel from the following recipe [9].

1) In a 500 mL beaker mix 25 mL of Silbond H-5 with 25 mL of ethanol 2) Then mix the catalyst solution in a 100 mL graduated cylinder; its components are 17.5 mL of ethanol, 37.5 mL of water, and 0.175 mL of 30% aqueous ammonia. The 0.175 mL of 30% aqueous ammonia requires a digital pipette due to the small amount of liquid. 3) Then pour the catalyst solution slowly into the Silbond H-5 solution stirring the entire time. This should take 45 seconds at minimum. 4) Then pour the resulting mixture carefully into the Teflon molds.

The Mold Stage

The gels should set in 30-90 minutes, assuming no temperature extremes [9]. After the gels are set, the following solution needs to be prepared [24].

17.5 mL ethanol. 37.5 mL water. Set the molds holding the samples in 50 mL beakers and completely cover them

with the new solution. This keeps the samples from being exposed to air which would dry out the samples and destroy the silicon lattice structure [24]. Cover the beakers with tin foil, or other material. Allow the gels to set in the molds for over 48 hours in order to give the lattice structure time to fully develop. The Teflon molds should not be tightened beyond snug due to their softness and the possibility of breakage.

The Bath Stage

After the gels are fully set, the next step is to remove them from the molds and set them in ethanol baths. Set aside a 50 mL beaker containing 25 mL of pure ethanol for each sample. Then, extract the samples from the molds by unscrewing the bottoms and gently pushing the gels into the prepared ethanol baths. The baths need to drained and replenished with an equal amount of ethanol every 8 to 24 hours. Each sample should be introduced to a total of at least 5 pure ethanol baths in order to ensure that the liquid held

54

within the gel matrix is pure ethanol. While able to be moved by hand, the more contact with pressure and air, the higher probability that the gel will weaken, develop cracks, or even break. The best method is to leave the gels in one beaker, and if they have to be moved, slide them gently from beaker to beaker. Even when care is taken in every step to minimize pressures and stresses endured by the gel, do not be surprised if about half the samples break, fall apart, or do not form a coherent gel matrix. The Drying Stage The Tousimis Samdri PVT-3D critical point dryer is used in the drying stage of this process. The critical point dryer chamber contains two openings, the liquid CO2 entry line and the exhaust line. Because the gels could potentially run into these openings and block up the lines, filter paper needs to be cut and folded into a cylinder with one end. Because it does not interact with the CO2 or ethanol, the filter paper will be able to effectively block the two openings from gel particulate while remaining invisible to the rest of the chemicals in the experiment. After the filter paper is in place, lining the drying chamber, pour ethanol into the chamber until it is half full. Then, extract the gel sample from its bath and place it into the drying chamber. Process the sample according to the Tousimis Advanced Manual Critical Point Dryer Samdri-PVT-3D user manual. The instructions for operation of the critical point dryer begin on page 15 and end on page 17 [27]. However, there is one important difference between the instructions and the procedures used in this project. As per Dr. A.J Tousimis’ suggestion, after step 7 is accomplished, the sample should be allowed to sit for 15-18 hours without moving on to step 8. After that time period, steps 6 and 7 are to be repeated and the gel needs to be allowed to sit in the same manner as before for 15-18 hours. During these wait times, the chamber does not need to be kept below ambient temperature. After the second waiting period, steps 6 and 7 are repeated for the third time and then the procedure continues as normal, with step 8 until the end.

When repeating steps 6 for the second and third time, it is imperative that viewing chamber be watched to guarantee that the liquid CO2 level does not drop below the sample. This misstep is easy to do because the liquid CO2 is cooling quickly from above ambient temperature to around 0 degrees Celsius. While cooling the pressure drops significantly and if it drops below 850 psi, a gas pocket develops in the chamber. If the liquid drops below the sample, allowing the gas pocket to come in contact with the sample then the sample quickly develops cracks and begins to turn opaque. This change in appearance represents the undermining of the solid matrix within the gel, and will result in shrunken, denser and highly-weakened samples. The critical point dryer, including the viewing chamber, can be seen in Appendix I.

Each time when performing step 6 from the Tousimis critical point dryer manual the beaker connected to the exhaust line should be monitored. When opaque white solids enter the beaker, looking like snow, then the critical point dryer chamber has been evacuated of ethanol and only liquid CO2 is left except for any ethanol involved in the gel matrix. This should only take about 10 minutes of purging and replacing the liquid in the chamber. An image of the exhaust beaker can be found in Appendix I. At the point when the CO2 is extracted for the final time, the gel sample becomes an aerogel.

55