Escolar Documentos
Profissional Documentos
Cultura Documentos
NEW RESEARCH
No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or
by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no
expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of information
contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in
rendering legal, medical or any other professional services.
ELECTROANALYTICAL CHEMISTRY:
NEW RESEARCH
GRAHAM M. SMITHE
Editor
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in
any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or
otherwise without the written permission of the Publisher.
For permission to use material from this book please contact us:
Telephone 631-231-7269; Fax 631-231-8175
Web Site: http://www.novapublishers.com
Independent verification should be sought for any data, advice or recommendations contained in this book. In
addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property
arising from any methods, products, instructions, ideas or otherwise contained in this publication.
This publication is designed to provide accurate and authoritative information with regard to the subject
matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering
legal or any other professional services. If legal or any other expert assistance is required, the services of a
competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED
BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF
PUBLISHERS.
Preface vii
Expert Commentary:
Electrodes Based on Metallophthalocyanines
Integrated with Carbon Nanotubes: Potential Hybrids
for Enhancing Electron Transport 1
Kenneth I. Ozoemena
Chapter 1 Clathrate Hydrate Crystallization for Clean
Energy and Environmental Technologies 9
Peter Englezos, John Ripmeester and Robin Susilo
Chapter 2 Corrosion Research Frontiers. Atmospheric Corrosion in
Tropical Climate. On the Concept of Time of Wetness
and Its Interaction with Contaminants Deposition 61
F. Corvo, T. Pérez, Y. Martin, J. Reyes,
L.R. Dzib, J.A. González and A. Castañeda
Chapter 3 The Application of D-Statistics Based Tests of Randomness,
Independence, and Trend to Electrochemical Observations 93
Thomas Z. Fahidy
Chapter 4 Self-Assembly Assisted Polypolymerization (SAAP):
A Novel Approach to Prepare Multiblock Copolymers
with a Controllable Chain Sequence and Block Length 109
Liangzhi Hong, Fangming Zhu, Guangzhao Zhang,
To Ngai and Chi Wu
Index 123
PREFACE
iscosahedral cavity with twelve pentagonal faces and six hexagonal faces (51268). It should be
noted that unlike structures I and II which may form with a single guest species, structure H
requires the presence of a small guest (like methane) and a large molecule guest substance
(LMGS) like neohexane at ordinary pressures.
Gas hydrates were first reported at the beginning of the 19th century, and until the 1930s
they remained a scientific curiosity. At that time it was realized that hydrates were more
likely to be the causative agent in blocking pipelines than ice. Today, gas hydrate control
continues to be a problem in the oil and gas industry. In the 1960’s it was realized that natural
gas hydrates are present in the geo-sphere with worldwide reserves estimated at 10,000 to
40,000 trillion cubic meters (TCM). Considerable efforts are underway to refine global
estimates and to develop technology and exploit this resource. On the other hand these
hydrates may decompose as a result of global warming or seafloor instability and release the
methane gas. There is speculation that a runaway greenhouse effect could result, with some
evidence for changes of such magnitude in the global paleoclimate (15,000 and 55 million
years ago). Application of clathrate hydrate crystallization offers the possibility of the
development of innovative technologies for natural gas storage and transportation, hydrogen
storage and gas separation with applications for carbon dioxide capture from flue gas
(CO2/N2/O2) or fuel gas (CO2/H2) mixtures. Clathrate hydrates have become an important
research area spanning a variety of disciplines. It is of continuing great interest to chemical
engineers who have played a key role in past developments. This report discusses areas where
chemical engineers can advance the knowledge frontier.
Chapter 2 - Atmospheric corrosion is the most extended type of corrosion in the World.
Over the years, several papers have been published in this subject; however, most of the
research has been made in non-tropical countries and under outdoor conditions. Results of
outdoor and indoor corrosion rate and corrosion aggressivity in tropical corrosion test stations
of Cuba and Mexico are reported.
Time of wetness (TOW), considered as the time during which the corrosion process
occurs, is an important parameter to study the atmospheric corrosion of metals. According to
ISO-9223 standard, TOW is approximately the time when relative humidity exceeds 80% and
temperature is higher than 0oC. No upper limit for temperature is established. In tropical
climates, when temperature reaches values over 25oC, evaporation of water plays an
important role and the possibility to establish an upper limit respecting temperature should be
analyzed.
The concept of TOW assumes the presence on the metallic surface of a water layer;
however, there are recent reports about the formation of water microdrops during the initial
periods of atmospheric corrosion, showing that the idea of the presence of thin uniform water
layers is not completely in agreement with the real situation in some cases (particularly indoor
exposures).
Most of the research carried out to study the initial stages of atmospheric corrosion have
been made on a clean surface without corrosion products; however, the metal is very often
covered by thin or thick corrosion products after a given exposure time and these products
usually act as retarders of the corrosion process.
In the Cuban Isle, the influence of chloride ions is very significant in determining the
corrosion rate. In the coastal territory of the Mexican Gulf, particularly at Campeche, the
deposition of Chloride ions is lower. No previous reports have been made about the
interaction between chloride deposition rate and rain. The influence of rain seems to be
Preface ix
important in determining the acceleration rate of chloride ions on metals due to its washing
effect. To consider the influence of the interaction chloride deposition rate–rain regime could
be useful to improve the prognosis of corrosion aggressivity. The predominant wind direction
corresponding to geographic sites result in an important parameter for chloride deposition and
their influence on surface wetness.
The calculation of Time of Wetness established in ISO 9223 should be revised based on
new results obtained in outdoor and indoor conditions in tropical humid marine climate.
Some proposals are made to improve the estimation of TOW, taking into account changes in
its nature depending on outdoor or indoor exposure, linear relationship between time and
TOW, the effect of rain, and the role of contaminants and air temperature.
Chapter 3 - Randomness, independence and trend (upward, or downward) are
fundamental concepts in a statistical analysis of observations. Distribution-free observations,
or observations with unknown probability distributions, require specific nonparametric
techniques, such as tests based on Spearman’s D – type statistics (i.e. D, D*, D**, Dk ) whose
application to various electrochemical data sets is herein described. The numerical
illustrations include surface phenomena, technology, production time-horizons, corrosion
inhibition and standard cell characteristics. The subject matter also demonstrates cross
fertilization of two major disciplines.
Chapter 4 - Block copolymers have attracted much attention because of their novel
properties and various promising potential applications. However, it is still difficult, if not
impossible, to prepare multiblock copolymers with a controllable chain sequence and block
length even though a variety of synthetic methods, such as anionic and controlled free radical
living polymerization have been advanced. In recent years, we have proposed and developed
a novel method of using the self-assembly of A-B-A triblock copolymers in a solvent which
is selectively good for the two A-blocks. Such self-assembly concentrates and exposes the
active groups attached on the two A-block ends so that they can be coupled together to form a
long multiblock copolymer chain with its sequence and block length controlled by the initial
triblock copolymer. In this review, we first illustrate how the SAAP concept was developed
and exemplified in some real copolymer systems. Furthermore, we compare the coupling
efficiency with and without the self-assembly, and demonstrate that SAAP provides an
elegant way to prepare long multiblock copolymers.
In: Electroanalytical Chemistry: New Research ISBN: 978-1-60456-347-4
Editor: G. M. Smithe © 2008 Nova Science Publishers, Inc.
Expert Commentary
ELECTRODES BASED
ON METALLOPHTHALOCYANINES INTEGRATED
WITH CARBON NANOTUBES: POTENTIAL HYBRIDS
FOR ENHANCING ELECTRON TRANSPORT
Kenneth I. Ozoemena*
Department of Chemistry, University of Pretoria, Pretoria 0002, South Africa.
Abstract
Carbon nanotubes (CNTs), notably single-walled (SWCNTs) or multi-walled (MWCNTs),
have continued to attract immense research interests in electroanalytical chemistry because of
their ability to exhibit unusual but excellent electrical conductivity and mechanical properties.
They also possess modifiable sidewalls and open-ends, making them suitable for use as new
materials for constructing efficient electrocatalysts and electrochemical sensors. Transition
metal metallophthalocyanine (MPc) complexes have also proved themselves as powerful
redox-active materials for modifying electrode surfaces for use as sensors. My group has been
engaged in the rational design and integration of CNTs with certain transition MPc
complexes, and exploring their potential applications in the fabrication of electrochemical
responsive CNT-MPc based electrodes. This short commentary gives insights into the recent
developments in this emerging research area as well as future trends.
1. Introduction
Phthalocyanine complexes are organic macrocycles with 18 π-electrons, structurally
resembling the naturally-occuring porphyrins complexes [1-3]. Electrodes modified with
transition metal (notably Fe, Co, Mn, Ni) phthalocyanine (MPc, Fig.1) complexes have
continued to generate immense research interests because of their well-established
electrocatalytic properties [3-6].
*
E-mail address: Kenneth.ozoemena@up.ac.za, Tel.: +27-12-420-2515; Fax: +27-12-420-4687 (Corresponding
author)
2 Kenneth I. Ozoemena
R R
Peripheral position
N
N N Non-peripheral position
N M N
N N
R R
In this method, about 5 µL of the CNT DMF solution (1mg / 1ml DMF) is first placed
onto the electrode and the solvent allowed to dry off in air or at mild oven temperature (ca. 80
o
C). The same process is repeated with DMF solution of MPc solution (1 mM). The
morphology of the films depends on the concentration of the casting solution, rate of solvent
evaporation, nature of the solvent and roughness of the electrode surface. Important substrates
for this are the pyrolytic graphites (highly oriented, basal or edge planes) because of the
inherent ability of these electrodes to interact with CNTs via π-π interactions [7,17,18].
Adhesion of CNT onto glassy carbon electrode (GCE) is difficult and fraught with problems
such as irreproducibility [19]. Abrasive immobilization is preferred for the MWCNT than the
more expensive SWCNT. The drop-coating is possible when the CNT is pretreated in harsh
acid conditions to introduce COOH, OH functionalities.
Self-assembly is a foremost strategy for forming highly stable, well-ordered ultra thin
‘self-assembled monolayer’ (SAM) films of redox-active species onto coinage metal surfaces
[20,21]. The use of SAMs for sensing and catalytic purposes is well documented. It involves
strong and irreversible chemisorption of the MPc complexes onto coinage metal based
electrodes, notably gold. Gold surfaces are preferred for forming thiol-derivatised SAMs
because of the well established specific and strong interaction of sulfur atoms with gold.
From past experiences in the formation of MPc-based SAMs [22-31], we (my research group
and collaborators) have now begun to exploit the SAM strategy in the integration of CNTs
with transition MPc complexes [32,33] (Figure 1). Amino-substituted metallophthalocyanine
complexes (MPc, Figure 2) can be covalently linked to acid-treated CNTs via the formation
of amide bond (if R = amino group [32]) or ester bond (if R = hydroxyl group [33]).
4 Kenneth I. Ozoemena
R R
N
N N Substituted MPc moiety
N M N
N N
N
R R
HO O OH
O
O O
Self-assembled carboxylated
SWCNT
O
O
O O
NH NH NH
NH
Self-assembled cysteamine
S S S S
Figure 2. Carton showing the integration of a substituted metallophthalocyanine (MPc) complex onto a
self-assembled single-walled carbon nanotube. The substitutent R could either be terminal hydroxyl (–
OH) group forming an ester bond, or amino (–NH2) group forming an amide bond.
t io n
e So lu
l y t
An a
MPc (re d uc e d )
MPc (o xidiz e d ) r id
c Hy b
-MP
CNT
Ele ctro de
e-
miniaturization and remote operation) over other analytical techniques such as the
spectroscopic and chromatographic methods, I envision that (i) important research works on
screen-printed electrodes for one-shot analysis, ultramicroelectrodes, lab-on-chips, microband
arrays, electrochemical sensors integrated with scanning electron microscopy and atomic
force microscopy, etc, are likely to dominate the development of electrochemical sensors in
the future; (ii) the prevalence of new types of diseases such as the drug-resistance
tuberculosis, and HIV and AIDS will continue to heighten the need for rapid and sensitive on-
site or point-of-care analysis; and (iii) the use of CNT-MPc composite (paste-based)
electrodes, especially with the less expensive MWCNTs than the more expensive SWCNT.
The main advantage of paste-based electrodes lies in their ease of regenerating the electrode
surface. For example, in situations where the electrode surface becomes irrevocably
contaminated or fouled new surface could easily be regenerated by polishing on clean
aluminum paper.
References
[1] Phthalocyanines: Properties and Applications, Lever, A.P.B.; Leznoff, C.C., Ed.; VCH
Publishers: New York, 1989, 1993, 1996, Vol.1–4.
[2] McKeown, N.B. Phthalocyanine Materials: Synthesis, Structure and Function,
Cambridge University Press: Cambridge, 1998.
[3] The Porphyrin Handbook, Kadish, K.M.; Smith K.M.; Guilard, R., Eds.; Academic
Press: Boston, 1999, Vol.1-10; and 2003, Vol. 11–20.
[4] Ozoemena, K.I.; Nyokong, T. In Encyclopedia of Sensors, Grimes, C. A.; Dickey, E.C.
Pishko, M.V., Eds.; American Scientific Publishers: California, 2006, Vol.3, Chapter E,
pp.157 – 200
[5] Nyokong, T. Coord. Chem. Rev. 2007, 251, 1707
[6] Vasudevan, P.; Phougat, N. Shukla, A.K. Appl. Organomet. Chem. 1996, 10, 591.
[7] Banks, C.E.; Moore, R.R.; Davies, T.J.; Compton, R.G. Chem. Commun. 2004, 1804
[8] Jurkschat, K.; Xiaobo, J.; Crossley, A.; Compton, R.G. Analyst 2007,132, 21.
[9] Banks, C.E.; Crossley, A.; Salter, C.; Wilkins, S.J.; Compton, R.G. Angew. Chem., Intl.
Ed. 2006, 45, 2533.
[10] Valcárcel, M.; Cárdenas, S.; Simonet, B.M. Anal. Chem. 2007, 79, 4788.
[11] Ozoemena, K.I.; Pillay, J.; Nyokong, T. Electrochem. Commun. 2006, 8, 1391.
[12] Pillay, J.; Ozoemena, K.I. Electrochim. Acta 2007, 52, 3630.
[13] Pillay, J.; Ozoemena, K.I. Chem. Phys. Lett. 2007, 441, 72-77.
[14] Siswana, M.; Ozoemena, K.I.; Nyokong, T. Electrochim. Acta 2006, 52, 114
[15] Silva, J. F.; Griveau, S.; Richard, C.; Zagal, J.H.; Bedioui, F. Electrochem. Commun.
2007, 9, 1629
[16] Liu, J.; Rinzler, A.G.; Dai, H.; Hanfer, J.H.; Bradley, R.K.; Boul, P.J.; Lu, A.; Iverson,
T.; Shelimov, K.; Huffman, C.B.; Macias, F.R.; Shon, Y.S.; Lee, T.R.; Colbert, D.T.;
Smalley, R.E. Science 1998, 280, 1253.
[17] Moore, R.R.; Banks, C.E.; Compton, R.G. Anal. Chem. 2004, 76, 2677.
[18] Wildgoose, G.G.; Leventis, H.G.; Streeter, I.; Lawrence, N.S.; Wilkins, S.J.; Jiang, L.;
Jones, T.G.J.; Compton, R.G. ChemPhysChem. 2004, 5, 669.
[19] Salimi, A.; Hallaj, R. Talanta 2005, 66, 967
Electrodes Based on Metallophthalocyanines Integrated with Carbon Nanotubes 7
Chapter 1
Abstract
Clathrate or gas hydrates are non-stoichiometric crystalline materials formed by the inclusion
of certain molecules into a framework of hydrogen-bonded water molecules under suitable
temperature and pressure conditions. The resulting host-guest networks consist of cavities
formed by water molecules enclosing the guest molecules. Typical guests include light
hydrocarbon gases, carbon dioxide, hydrogen sulfide, hydrogen and nitrogen. The basic cavity
formed by water molecules through hydrogen bonding is the pentagonal dodecahedron (512).
This cavity is common to the three best known hydrate structures (I, II and H). A unit cell of
structure I hydrate has 46 water molecules forming two dodecahedral (512) and six
tetrakaidecahedral (51262) cavities. A unit cell of structure II has 136 water molecules forming
16 (512) and eight hexakaidecahedral (51264) cavities. The Structure H hydrate unit cell has 34
water molecules, three 512 cavities, two different dodecahedral cavities which have three-
square faces, six-pentagonal faces and three-hexagonal faces (435663) and a larger iscosahedral
cavity with twelve pentagonal faces and six hexagonal faces (51268). It should be noted that
unlike structures I and II which may form with a single guest species, structure H requires the
presence of a small guest (like methane) and a large molecule guest substance (LMGS) like
neohexane at ordinary pressures.
Gas hydrates were first reported at the beginning of the 19th century, and until the 1930s they
remained a scientific curiosity. At that time it was realized that hydrates were more likely to
be the causative agent in blocking pipelines than ice. Today, gas hydrate control continues to
be a problem in the oil and gas industry. In the 1960’s it was realized that natural gas hydrates
are present in the geo-sphere with worldwide reserves estimated at 10,000 to 40,000 trillion
cubic meters (TCM). Considerable efforts are underway to refine global estimates and to
develop technology and exploit this resource. On the other hand these hydrates may
10 Peter Englezos, John Ripmeester and Robin Susilo
decompose as a result of global warming or seafloor instability and release the methane gas.
There is speculation that a runaway greenhouse effect could result, with some evidence for
changes of such magnitude in the global paleoclimate (15,000 and 55 million years ago).
Application of clathrate hydrate crystallization offers the possibility of the development of
innovative technologies for natural gas storage and transportation, hydrogen storage and gas
separation with applications for carbon dioxide capture from flue gas (CO2/N2/O2) or fuel gas
(CO2/H2) mixtures. Clathrate hydrates have become an important research area spanning a
variety of disciplines. It is of continuing great interest to chemical engineers who have played
a key role in past developments. This report discusses areas where chemical engineers can
advance the knowledge frontier.
Introduction
It is well known that when sufficient amounts of water and a hydrate-forming substance
are brought into contact under appropriate temperature and pressure conditions a crystalline
solid known as a gas or clathrate hydrate forms (van der Waals and Platteeuw, 1959;
Davidson, 1973; Englezos, 1993; Sloan, 1998, Ripmeester, 2000; Koh, 2002; Sloan, 2003a;
2003b; 2004a; 2004b; 2005; Englezos and Lee, 2005; Chatti et al. 2005; Bishnoi and Clark,
2006). Gas hydrates were reported as early as 1810 (Davy, 1811). Following Sir Humphrey
Davy’s report of aqueous solutions of chlorine that remained solid at temperature above 0°C
(Davy, 1811) and Faraday’s confirmation in 1823, gas hydrates were a steady object of
scientific curiosity for more than 100 years. Research efforts became more focused from the
mid 1930’s on due to the suggestion that the unwanted solid material in gas transmission
pipelines was in fact gas hydrate rather than ice, frequently forming plugs at temperatures
above the icepoint (Hammerschmidt, 1934). Since then, extensive experimental and
computational studies have been carried out in order to identify the equilibrium formation
conditions for various hydrate forming systems. This research was driven by the need to
establish methods to prevent the occurrence of hydrate of hydrate plugs in oil and gas
pipelines. Hydrate research on flow assurance is still carried out especially with oil and gas
exploration into deeper water and remote offshore areas. Despite the initial negative
impression on gas hydrates, recently it has been realized that gas hydrates possess important
roles towards energy and environmental issues as well, mainly due to its potential for gas
holding capacity and separation purposes.
Methane, and other natural gas components, trapped in ice-like lattices known as ‘gas
hydrates’ or ‘clathrate hydrates’ have been found to exist naturally in the earth, especially
offshore on the continental margins and under the permafrost in the Arctic (Makogon et al.,
1972; 1987; Suess et al., 1999; Kvenvolden 1988; 1999; 2000; Reeburgh, 2003; Buffet and
Archer, 2004; Klauda and Sandler, 2005). The exact quantity of natural gas in hydrate form is
not known accurately, but is significant and considered to be a huge potential unconventional
energy source for the future, with the latest estimates some 5 – 20 % of the global carbon
budget. Unfortunately technologies for gas production do not exist as yet due to
complications involved. Besides the economic aspect that has to be taken into consideration,
the impacts on the environment, ecology, and geological stability are also important. The
decomposition of natural hydrate may destabilize and alter the earth’s geological features
(Glasby, 2003; Sultan et al., 2004a; 2004b) causing geohazards such as landslides,
earthquakes, and even tsunamis if the mass movements are under water. On the other hand,
current global warming may initiate the decomposition of natural hydrates (Hatzikiriakos and
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 11
Englezos, 1993; Kvenvolden, 2002). The first production test of methane hydrate was
conducted in the Mackenzie Delta in the Canadian Arctic. Two test wells were drilled where
samples cores were collected to understand its characteristic and evaluate the possibility for
successful gas production (Bybee, 2004). Thermal and pressure stimulations were tested to
measure both input conditions and reservoir responses that enabled the calibration and
refinement of reservoir simulation models (Collet, 2005). Moreover, hydrates were also
suggested to exist in extraterrestrial space especially on the outer planets and satellites, eg.
Mars, Saturn, Uranus and Neptune (Delsemme and Swings, 1952; Delsemme and Wenger,
1970; Lunine and Stevenson, 1985; Koh, 2002; Osegovic and Max, 2005; Tobie et al., 2006;
Machida et al., 2006; Hand et al., 2006). Titan, the largest moon of the Saturn is believed to
have 100 km thickness of high pressure methane hydrate within its ice mantle (Loveday et al.,
2001).
Gas hydrate also offers opportunities to develop innovative technologies for gas storage
and separation (Gudmundsson et al., 1994; 2000; Mori, Y. H., 2003; Englezos and Lee, 2005)
as well as cool energy storage (Inaba, 2000; Tanasawa and Takao, 2002). The applications
include storing and transporting natural gas (Khokkar et al., 1998; Seo and Lee, 2003; Mori,
2003; Thomas and Dawe, 2003; Lee et al., 2005b; Tsuji et al., 2005a; 2005b; 2004;
Javanmardi et al., 2005; Abdalla and Abdullatef, 2005), hydrogen storage for the hydrogen
economy of the future (Mao et al., 2002; Patchkovskii and Tse, 2003; Lee et al., 2005a;
Schuth, 2005 Strobel et al., 2006; Hester et al., 2006; Hu and Ruckenstein, 2006),
sequestration of carbon dioxide with in situ methane hydrate decomposition (Lee et al. 2003;
Park et al., 2006a; House et al., 2006), separation of carbon dioxide from the flue gas (Seo et
al., 2005a; Yoon et al., 2006; Park et al., 2006b), seawater desalination (Parker, 1942;
Barduhn et al., 1962; Javanmardi and Moshfeghian, 2003; Rautenbach and Pennings, 1973)
and refrigeration (Tomlison, 1982; Ternes, 1984; Mori and Mori, 1989a,b; Bi et al., 2006).
For hydrate technology to be industrially applicable, several challenges need to be addressed
such as: high gas storage capacity, efficient gas separation, fast phase transformations, and
mild pressure-temperature conditions during processing, storage, and transportation.
Research on gas hydrates involves engineers and scientists. Chemical engineers in
particular can make significant contributions towards the commercialization of gas hydrate-
based technologies. Therefore, the objective of this chapter is to review developments and
highlight areas of opportunity for the involvement of chemical engineers. The structure of the
chapter is as follows. First clathrate hydrates are described with emphasis on their structures
and the fundamentals of thermodynamic and kinetic properties. Subsequently, the clean
energy technologies under development are discussed. Here, chemical engineers can play a
crucial role in the successful scale-up of the various existing concepts. These include natural
gas storage and transport, hydrogen storage, and carbon dioxide capture and sequestration.
Flow assurance in the hydrocarbon production and transportation industry is the only
industrial-scale area where continuous research activities both at fundamental and practical
levels are ongoing. However the potentials of hydrates for other practical applications have
not been exploited mostly because the technology is relatively new and not established.
Finally the recovery of natural gas from natural deposits is briefly discussed and the interested
reader is directed to a number of recent references.
12 Peter Englezos, John Ripmeester and Robin Susilo
Clathrate Hydrates
Gas hydrates are true inclusion compounds and hence also are known as clathrates
(Davidson, 1973). They are non-stoichiometric crystalline solids with the physical appearance
of ice, and consist of water molecules that serve as the host material and guest molecules
trapped in the host. Natural gas components such as: methane, ethane, propane, carbon
dioxide, and hydrogen sulfide are typical guest molecules, although any hydrophobic
molecule that fits the host cavities can be the guest. The water molecules are connected
through hydrogen bonds forming cages/cavities that completely encage individual guest
molecules. The crystal lattice consisting of empty cages is not stable thermodynamically, and
there is a minimum guest content required to give a stable lattice. Hence, guest molecules
must have the correct sizes and geometry in order to fit into the different cages. Normally
there is a single molecule per cage except in the case of high pressure hydrates of atoms or
small molecules (N2, H2, CH4, rare gases). There are no specific or directional interactions
between the host and guest molecules with the weak van der Waals forces providing the key
interaction. Thus the guest molecule is free to rotate and translate inside the cage. The crystal
lattices have limited flexibility to accommodate guests of different size so that the crystal
lattice parameters have well defined limits. Further discussion of structures, kinetics and
thermodynamic properties are discussed in the next section.
Structures
Hydrate cages are arranged and packed in different configurations, with three distinct
crystal structures commonly encountered today: cubic structure I (sI) (Müller and
vStackelberg, 1952; McMullan and Jeffrey, 1965), cubic structure II (sII) (vStackelberg and
Müller, 1951; Mak and McMullan, 1965), and hexagonal structure H (sH) (Ripmeester et al.,
1987). The cubic structures have two types of cavities (small/S and large/L) but the hexagonal
Structure I II H
Crystal system Cubic Cubic Hexagonal
Space group Pm3n Fd3m P6/mmm
a = 12.2 A
Lattice parameters a = 12A a = 17.3A
c = 10A
Number of cage 2 2 3
Small/S (512)
Small/S (512) Small/S (512) Medium/M
Cage identification
Large/L (51262) Large/L (51264) (435663)
Large/L (51268)
Ideal unit cell formula 2S.6L.46H2O 16S.8L.136H2O 3S.2M.1L.34H2O
rS = 3.91A
rS = 3.95A rS = 3.91A
Cage radius rM = 4.06A
rL = 4.33A rL = 4.73A
rL = 5.71A
S = 20
S = 20 S = 20
Coordination number M = 20
L = 24 L = 28
L = 36
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 13
one has three cavities (small/S, medium/M and large/L). Experience has shown that the large
cavities in all common hydrate structures usually are filled completely by a guest molecule.
For the cubic structures, the small cavity may be empty and this depends on molecular size,
so that a single guest species is enough to stabilize sI and sII hydrates. However a small “help
guest” is required to maintain sH hydrate so that two guest species are required. Generally
there is only one guest molecule in a cage. Few exceptions were reported at high pressures
where smaller molecules like nitrogen (Kuhs et al., 1997), hydrogen (Mao et al., 2002), rare
gases (Loveday et al., 2003a; Alavi et al., 2005; 2006b), and methane (Loveday et al., 2001;
2003a) may fill and stabilize the large cage of sII and sH with multiple guest occupancies of a
cage.
- Structure I (sI)
- Structure II (sII)
Structure H (sH)
are located on the equatorial plane and linked to another large cage via the medium cages
(435663). The other two-hexagonal faces are located at the polar plane and connected to
another large cage. The small cage (512) fills the space remaining in the unit cell. The cages
are shown in Figure 1 and their properties are summarized in Table 1. sI and sH hydrate both
can be derived from the stacking of layers of pentagonal dodecahedra. The space between
layers stacked in different ways is taken up by the other cages in the crystal lattice.
The stable crystal structure formed depends on the guest molecule(s) present. Guest
molecules that fill the void space of the cages efficiently are generally preferred. As expected,
see Table 1, the size of all of the dodecahedral cages in the hydrate structures is quite similar
although the cage symmetry is different in the three structures. However the large cage size
increases on going from sI, to sII and sH. Hence the stable hydrate structure formed is
governed by the size of the largest molecule(s) in a particular hydrate. Guest molecules with a
van der Waals diameter less than ~9.7A have been reported as hydrate formers. Structure I
hydrate is generally formed by smaller molecules (4.2A<d<6A), sII with intermediate sized
molecules (6A<d<7A) and sH with larger molecules (7A<d<9.7A) (Sloan, 1998; Ripmeester
et al., 1994; Ripmeester and Ratcliffe, 1990). Long chain hydrocarbons with more than 6
carbon atoms or longer than 10A are considered as non-hydrate former. Recently n-pentane
and n-hexane were discovered as sH hydrate formers, too, in the presence of a good sH
hydrate former such as 2,2-dimetyl butane (Lee et al., 2006a). Smaller molecules with a van
der Waals diameter less than ~4.2A generally form sII hydrates.
Many phase properties of solid hydrates can be determined by instrumentation suitable
for molecular-scale characterization techniques such as X-Ray diffraction, neutron
diffraction, Raman and Nuclear Magnetic Resonance (NMR) Spectroscopy. The diffraction
techniques allow the determination of the solid phase structures. Powder X-Ray diffraction
(PXRD) measurements provide information on the crystal dimension and lattice parameters
only (Takeya et al., 2006). More complete information regarding the atomic coordinates and
occupancies is obtainable from single crystal X-Ray diffraction (Udachin et al., 2002;
Kirchner et al., 2004), although powder methods for deriving more complete structural
information are under continuous development (Chazallon and Kuhs, 2002; Aladko et al.,
2006b). Structural information can also be inferred from Raman spectra (Sum et al., 1997;
Tulk et al., 2000) and NMR chemical shifts and dynamically averaged lineshapes (Ripmeester
and Ratcliffe, 1990; Ripmeester and Ratcliffe, 1999). The NMR spectra generally provide
structure-specific signatures for each guest in the different hydrate cages of different size and
shape. The intensity of the signals from Raman and NMR spectroscopy may also provide
information on the cage occupancies (Tulk et al., 2000; Wilson et al., 2002), although the
NMR methods provide the only truly quantitative analyses. Raman spectroscopy requires
calibration in each specific case, so NMR is more suitable for determining cage occupancies.
The thermodynamics of gas hydrates is a mature subject (Sloan, 2004b). However, new
data continue to appear, mainly for mixtures required in new applications, or for more severe
conditions of temperature and pressure. A list with experimental data up until the late 1990s is
available (Sloan, 1998). From a computational standpoint the main interest is in calculating
the incipient hydrate formation pressure for a given hydrate-forming mixture at a particular
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 15
temperature. Modeling of hydrate phase equilibria has been based on the statistical
thermodynamics model of van der Waals and Platteeuw (van der Waals and Platteeuw, 1959;
Parrish and Prausnitz, 1972) and various computer methods have been developed.
Hammerschmidt developed the first method used in the industry for predicting the inhibiting
effect of methanol (Hammerschmidt, 1934). The method is empirical and the reliability of the
calculations is variable (Ng and Robinson, 1985). Anderson and Prausnitz presented a
thermodynamics-based method for calculating the inhibiting effects of methanol (Anderson
and Prausnitz, 1986). They used the van der Waals-Platteeuw model for the solid hydrate
phase, Redlich-Kwong equation of state for the vapour phase and the UNIQUAC model for
the liquid phase. Henry’s constants were used for calculating the fugacities of components in
their supercritical state in the liquid phase. Furthermore, empirical correlations were used for
calculating the molar volumes, partial molar volumes at infinite dilution and the fugacity of
the hypothetical liquid phase below the ice-point temperature. Robinson and Ng presented a
computer program for the calculation of the depression of hydrate formation temperatures due
to methanol (Robinson and Ng, 1986).
A computational method based on the Trebble-Bishnoi equation of state was presented by
Englezos et al. for calculating the depression effects of methanol and the amounts of
methanol required (Englezos et al., 1991; Trebble and Bishnoi, 1988). All equilibrium hydrate
prediction methods use either an equation of state for all fluid phases or an equation of state
for the vapor phase and activity coefficient models for the liquid (Englezos, 1993). The use of
a molecular-based equation of state (SAFT) to compute the inhibiting effect of methanol,
glycols and glycerol was also presented (Li and Englezos, 2006; Li et al., 2007). More
recently Sloan’s group at the Colorado School of Mines (CSM) provided software to predict
the hydrate phase equilibria for mixtures and in the presence of thermodynamics inhibitors
(Ballard and Sloan, 2002; Jager et al., 2003; 2005; Ballard and Sloan, 2004a; 2004b). Flash
calculations in which one is also interested in determining the amount of hydrate formed have
also appeared (Bishnoi et al., 1989). It is difficult to determine the quantity of hydrate or
hydrate phase fraction experimentally, and this is an area in need of future attention.
Kinetics is concerned with the rate at which the phase transformation occurs and the
identification of the factors affecting it. Gas hydrate formation, being a crystallization process
is characterized by nucleation followed by crystal growth and agglomeration (Bishnoi and
Natarajan, 1996; Englezos, 1996). The rate of nucleation e.g. number of hydrate crystal nuclei
formed per unit time per unit volume is extremely difficult to measure and to date there is no
data available. Most gas hydrate kinetic studies have focused on the growth phase and involve
measuring the rate of uptake of the hydrate forming substance (Englezos et al., 1987a;
Englezos et al., 1987b). This is a macroscopic measurement which may also be coupled with
measurement of the particle size distribution (Clarke and Bishnoi, 2004; Clarke and Bishnoi,
2005). Bishnoi’s laboratory at the University of Calgary pioneered these kinetic studies. They
also develop a mechanistic hydrate kinetics model (Englezos et al., 1987a; 1987b) that was
later improved (Dholabai et al., 1993). Another macroscopic-type measurement coming
online is based on differential scanning calorimetry (Le Parlouer et al., 2004).
Figure 2 shows the typical methane uptake profile (Lee et al., 2005b). An initial increase
in methane consumption is observed due to the gas dissolution in water. Once the liquid is
saturated with the gas, no more gas consumption is seen. The hydrate nuclei start to form at
this point until it reaches a critical size as indicated by the arrow and an increase in the
temperature profile due to heat released during the hydrate formation (exothermic). The curve
16 Peter Englezos, John Ripmeester and Robin Susilo
representing methane moles consumed versus time is approximately linear after the
nucleation but the slope of the curve gradually decreases at longer experimental times after a
certain level of conversion to hydrate. The decrease was suggested as being due to limited
contact between the water and gas.
0.20 277.5
Temperature (K)
276.5
0.10
276.0
0.05 275.5
275.0
0.00
ntb (0.0071mol, 482.7 min)
274.5
0 100 200 300 400 500 600 700 800
Time (min)
Figure 2. Rate of methane consumption for sI hydrate formation of CH4-H2O system at 275.50 K and
3.96 MPa. (Reprinted from Energy and Fuels (Lee et al., 2005b), Copy right (2005) with permission
from American Chemical Society).
The macroscopic approach does not provide any information on the solid phase
properties (Sloan, 2003b). Recently a more sophisticated in-situ measurements using Raman
spectroscopy (Park et al., 2006a), NMR spectroscopy and imaging (Moudrakovski et al.,
2001; 2004; Gao et al., 2005; Seo et al., 2005a; Park et al., 2006a; Chen et al., 2006; Susilo et
al., 2006), X-Ray diffraction (Uchida et al., 2003; Takeya et al., 2002), neutron diffraction
(Kuhs et al., 2006; Hester et al., 2006) and digital scanning calorimetry (Dalmazzone et al.,
2006) have been employed to follow hydrate conversion. Molecular dynamics simulation is
also being exploited to understand the kinetics of hydrate formation (Nada, 2006; Vatamanu
and Kusalik, 2006).
In order to gain more insight into hydrate kinetics, hydrate crystal morphology has also
been studied in order to elucidate the mechanism of hydrate nucleation, migration, and
growth at the surface of water droplets and at the interface between water and gas or non-
aqueous liquid (Lee et al., 2005c; 2006b; Ohmura et al., 2004; 2005b; 2005d; Servio and
Englezos, 2003a; 2003b; Ito et al., 2003; Kobayashi et al., 2001; Uchida et al., 1999b). A
hydrate film growth model can be derived from the macroscopic observation (Mochizuki and
Mori, 2006; Mori, 2001; Mochizuki, 2003). In general, hydrate preferentially nucleates at an
interface where the guest concentration is at a maximum. However hydrate nucleation from
the bulk liquid phase was also observed and this occurs only when the liquid phase is
supersaturated or where there is local sub-cooling. The nucleation and crystal growth rates
generally increase with driving force, given by pressure increase or temperature decrease
from the equilibrium values. After nucleation, hydrates tend to grow into the adjacent
volumes. If the nucleation occurs at an interface, hydrates will grow along the interface,
forming a film with a certain thickness. If the nucleation occurs in the bulk phase, the crystal
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 17
will grow and then float towards the interface if the density of the hydrate is lower (e.g.
methane hydrate) than that of water. Subsequent hydrate growth is limited by mass transfer of
water or guest molecules across the hydrate film. It has been suggested that the gas may
diffuse and grow towards the water phase and water may also migrate out through the hydrate
due to capillary action. Hairy or needle-like crystal morphologies were observed as going
from the interface towards the gas phase. Columnar or dendritic crystals were observed as
growing from the interface towards the liquid water phase. Crystal morphology has been
suggested to depend on the degree of sub-cooling (Knight and Rider, 2002; Lee et al., 2006b).
The presence of inhibitors changes the crystal morphology (Lee and Englezos, 2006; Zeng et
al., 2003, Larsen et al., 1998).
Kinetics through gas uptake measurements and macroscopic techniques, in general, is in
reality the “average kinetics” over the whole sample. The observation of gradual conversion
in bulk samples only arises as a result of averaging over many local environments
(Moudrakovski et al., 2004), as locally the conversion to hydrate is quite an inhomogeneous
process. This significant finding, based on microscopic techniques such as NMR
microimaging, poses the question how microscopic measurements of kinetics are related to
kinetics obtained from macroscopic techniques. Susilo et al. (2006) employed Nuclear
Magnetic Resonance (NMR) spectroscopy and imaging (MRI) to monitor the kinetics of
structure I and H methane hydrate growth and found that the results agree with the results
from gas uptake measurements obtained by Lee et al. (2005a). At least, both methods give the
same answers when the local techniques are averaged over the sample inhomogeneities.
Kinetic studies involve measurement of induction times for crystallization and the
determination of the rate of hydrate crystal growth, and the latter is usually defined
operationally. Thus, one may use the gas uptake rate to describe kinetics or the rate at which
an interface involving a hydrate phase advances. The question that arises is how an intrinsic
rate can be distinguished from the relevant transport processes. The difficulty in this regard as
well as the variety of methods to study the dynamics of hydrates gives rise to a variety of
kinetic studies on gas hydrates. In addition, in spite of kinetic models that have appeared and
aided the mechanistic understanding of hydrate growth it is still difficult to come up with a
predictive model that will allow prediction of the onset of hydrate formation. Thus, there is a
need to study particular hydrate forming systems and hydrate vessel configurations. This is an
area for future growth and it is ripe for new approaches. In this regards, a new computer
simulation approach based on cellular automata and Monte Carlo methods appeared recently
(Kvamme et al., 2004; Svandal et al., 2006; Buanes et al., 2006).
It is also important to mention the nucleation time during hydrate formation. Nucleation
by far is known as a stochastic phenomenon hence it is unpredictable meaning the induction
data cannot be reproducible from system to system even though each system has the same
thermal history (Lingelem et al., 1994). The time required to nucleate hydrate seed may
further delay the productivity of hydrate formation. It is suggested that heterogeneous
nucleation occurs in most hydrate systems due to lower driving force (higher temperature or
lower pressure) required than homogeneous nucleation (Zeng et al., 2006a,2006b). Hydrate
nucleation time is also affected by the history of water sample but the growth is not. Liquid
water and also ice that has experienced hydrate formation tends to have a shorter nucleation
time (Moudrakovski et al., 2001; Buchanan et al., 2005; Lee et al., 2005c). This is also
referred as memory effect. It is still unclear and debatable the reasoning behind the “memory
effect”. It was suggested that there is residual hydrate structures and/or dissolved gas in
18 Peter Englezos, John Ripmeester and Robin Susilo
memory solution. Neutron diffraction study did not detect any residual hydrate structures in
the memory solution as previously suggested (Buchanan et al., 2005). The structure of water
changes only when hydrate crystallites are present in the solution (Thompson et al., 2006).
Perhaps the dissolved gas in liquid water after hydrate decomposition alters the water
molecule orientation due to hydrophobic interaction that leads to faster hydrate reformation.
The statistical study on induction time indeed show that the “memory effect” tends to
disappear when the samples are exposed to higher temperature or longer decomposition time
(Ohmura et al., 2003b). Nevertheless, the utilization of memory solution is beneficial to speed
up the nucleation time. This is an important aspect for industrial practice. The water from
dissociated hydrate can be sent back to the hydrate production plant where it is reused to
speed up hydrate nucleation. Anti-freeze protein (AFP) is the only inhibitor that that has
shown to have the ability to successfully eliminate the memory effect (Zeng et al., 2006a;
2006b). a combination of a commercial inhibitor and polyethylene oxide was found to
substantially suppress the memory effect (Lee and Englezos, 2006).
Molecule ΔH H-L-V
Hydrate former Structure P equilibria [MPa]
size [A] [kJ/mol]
Hydrogen 2.72 II [152] >100 [151] NA
Nitrogen 4.10 II [106] 16.10 [311] 49.54 [41*]
Oxygen 4.20 II [275] 12.02 [310] 49.16 [41*]
Methane 4.36 I 2.65 [95] 54.19 [71]
Hydrogen sulfide 4.58 I 0.10 [235] 63.10 [321]
Carbon dioxide 5.12 I 1.22 [314] 65.22 [98]
Ethane 5.50 I 0.50 [45] 71.80 [71]
Tetrahydrofuran 5.90 II 0.1; 277K [70] 99.50 [295]
Propane 6.28 II 0.16 [174] 129.2 [71]
i-butane 6.50 II 0.11 [231] 133.2 [71]
Methyl tert-butyl ether 7.60 H 1.52 [84;*] 76.6 [**]
Neohexane 7.99 H 1.13 [166,173,200,*] 76.8 [**]
1,1-dimethylcyclohexane 8.40 H 0.88 [45;*] 81.8 [**]
Methylcyclohexane 8.59 H 1.29 [173,177,200;*] 79.5 [**]
*
Phase equilibria - methane occupies the 512 and 435663 cages
**
Regressed using Clausius-Clayperon equation from the H-L-V phase line
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 19
molecule size is correlated with hydrate structures with exception from those of the smaller
ones. Hydrate decomposes below the equilibrium pressure at a given temperature. Higher
pressure is required at higher temperature. The decomposition process is an endothermic
reaction, which absorb heat that can be measured by calorimetric method or regressed from
the corresponding phase equilibrium data (Yoon et al., 2003; Sloan and Fleyfel; 1992).
Hydrogen is the simplest molecule and it has been suggested to be the starting material
for the creation of planets (Stevenson, 1999). Hydrogen is also a suitable fuel for fuel cells.
Nitrogen and oxygen are the essential components of air which have been found trapped
within the polar ice sheet in the form of bubbles as well as hydrates. The explanation of how
air hydrate forms in the deep glaciers such as in Antarctica and the hydrate composition may
help explain some of the paleoclimate records. (Uchida et al., 1994). Most of the other
hydrate formers listed in Table 2 are typical components of natural gas that are commonly
found in gas reservoirs or natural hydrate samples. Tetrahydrofuran is a water soluble guest
that is known to have a strong hydrate stabilizing effect. It is particularly attractive for
hydrogen storage where a tremendous pressure reduction can be achieved by forming mixed
hydrogen and THF hydrate (Florusse et al., 2004; Lee et al., 2005a; Alavi et al., 2006a). The
large molecules that occupy the large cage of sH hydrate are also called LMGS (Ohmura et
al., 2002; Tsuji et al., 2004). LMGS may be found as condensates in gas transmission
pipelines and in naturally occurring sH hydrate (Lu et al., 2007). The LMGS in condensate
are undesirable from the point of view of flow assurance because methane and LMGS may
form hydrate and plug the pipelines. On the other hand, the LMGS may be beneficial for gas
storage and transport applications as methane can be stored under milder conditions. The
maximum pressure reduction for methane storage in sH hydrate was reported for 1,1-
dimethylcyclohexane (Hara et al., 2005). The following section discusses the opportunities
for gas hydrate applications to clean energy and the environment. The motivation, current
knowledge, opportunities, challenges, research efforts in our laboratories and directions
towards commercializing the technology are discussed.
be fractionation of the components between the hydrate and the hydrocarbon phase. The
formation of the thermodynamically favored phases should result in the lowest equilibrium
pressure. If the natural gas contains only methane, the stable hydrate structure is sI. However
the most stable hydrate structure is more likely to be sII due to the presence of larger
components that are sII hydrate former such as propane, i-butane (Østergaard et al., 2001),
and various combinations of ethane and methane also give sII hydrate (Hester and Sloan,
2005; Takeya et al., 2003; Subramanian et al., 2000a; 2000b). The formation of sH hydrate is
possible if the gas contains condensates. A mixture of hydrate structures may also be present.
Methane gas is the preferred model system representing natural gas in most hydrate study
because it is the cleanest fuel among hydrocarbons. The simplified system provides a sound
basis for further development and avoids the complexities that may be encountered when
working with gas mixtures (Schicks et al., 2006) although the ideal methane system may not
approximate real life. Produced natural gas also contains unwanted compounds such as
hydrogen sulfide, carbon dioxide, and heavier hydrocarbon. Thus one has to treat the gas first
to meet the safety regulations before converting it into hydrate. The fractionation of produced
natural gas could also be performed with hydrate technology. Hydrate fractionation allows the
gas phase to be methane-rich due to the preference of the heavier hydrocarbon components as
well as CO2 and H2S for the hydrate phase. This enables the trapping of methane in hydrate
after fractionation (Uchida et al., 2004). Obviously hydrate processes have to be coupled to
other processes such as membrane technology to further purify the gas phase.
Methane is one of the most studied guest molecules associated with gas hydrate. This is
simply because methane is found in huge quantities in natural gas, coalbed methane, and
natural hydrates, and it is the cleanest fuel among hydrocarbons. Under moderate pressure
conditions, methane and water form a stable sI hydrate with the large and most small cages
occupied by methane if the hydrate is prepared near equilibrium conditions. Hydrate phase
equilibrium data, dating back to 1946, has been reported by many researchers (Deaton and
Frost, 1946). Hydrate phase equilibrium data at temperature below the icepoint is also
available (Makogon and Sloan, 1994). The phase diagram can be predicted satisfactorily by
using the van der Waals-Platteew model (Van der Waals and Platteeuw, 1959). The model
was improved by Sloan’s group at the Colorado School of mines to better predict the hydrate
phase equilibria for mixtures and in the presence of thermodynamics inhibitors (Ballard and
Sloan, 2002; Jager et al., 2003; 2005; Ballard and Sloan, 2004a; 2004b). The hydration
number of methane sI hydrate has been measured by various techniques such as calorimetry
(Handa, 1986a,b), Raman (Uchida et al., 1999a; 2003; Sum et al., 1997) and NMR
spectroscopy (Collins et al, 1990; Ripmeester et al., 1988) with the value reported quite
consistent at around ~6.0 ±0.2 irrespective whether it is a synthesized or natural hydrate
formed at different conditions. Methane cage occupancies are also given in the literature
obtained from Raman (Sum et al., 2000), NMR (Ripmeester et al., 1988), and single crystal
X-Ray diffraction (Udachin et al., 2002; Kirchner et al., 2004). They are generally in good
agreement with the predictions of the statistical thermodynamic model, which suggests that
all large cages are fully occupied with methane and approximately 90% of the small cages are
occupied when hydrates are prepared near Hydrate-Gas-Liquid water (HG-L) equilibrium
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 21
conditions. The measured and calculated hydration numbers decrease slightly with increasing
pressure with higher methane cage occupancy (Sum et al., 1997; Klapproth et al., 2003;
Circone et al., 2005). The reported hydration number at very high pressure (620 MPa) is 5.67
(Ogienko et al., 2006), which correspond to full methane occupancy of both cages. The lattice
constant of sI methane hydrate for synthetic and natural hydrate is practically the same at
11.85A (113K) which corresponds to hydrate density of 0.95 g/cm3 (Takeya et al., 2006). The
lattice constant increases with temperature due to thermal expansion (Shpakov et al., 1998)
but decreases with pressure due to compressibility (Klapproth et al., 2003). The reported
lattice constant at 273K is 11.99A and refers to a hydrate density of 0.92 g/cm3 (Takeya et al.,
2006). Consequently the methane content in hydrates can be calculated based on the
hydration number and density, which is approximately ~160-180 v/v depending on the cage
occupancy and temperature.
Gas storage capacity in hydrate is one of the most important factors and depends strongly
on the hydrate structure and its ability to accommodate the guest molecules. Each hydrate
structure has its characteristic cage sizes and cage packing configuration. Methane is a small
molecule with a van der Waals diameter of 4.36A and can fit into all of the hydrate cages.
However there is a preference regarding the cages over which methane is distributed as driven
by thermodynamic and kinetic factors. Methane and water generally form a stable sI hydrate
under moderate pressures. The formation of stable pure methane sII or sH hydrates occurs
only at extremely high pressure (25 MPa or above) (Chou et al., 2000). In the presence of
another guest molecule such as double hydrates are formed that are stable eg propane for sII
(Kini et al., 2004) and methylcyclohexane for sH (Mehta and Sloan, 1993). A sII methane
hydrate was also reported as a kinetic product (meta-stable) (Schicks and Ripmeester, 2002).
Methane itself does not stabilize the sII or sH lattices at moderate pressures because of
inefficient interactions between the host and guest, and hence the presence of a larger
molecule is required. The methane content of all hydrate structures is summarized in Table 3.
The number for LNG and the corresponding hydrate stability conditions are also given. It is
again important to note that not all hydrate cages are fully occupied by methane under three
phase equilibrium conditions. The numbers for full methane occupancy and the realizable
methane content are also given. For sII and sH hydrates most or all of the large cages are
filled with a larger guest molecule.
It is clear from Table 3 that storing methane in LNG has the maximum energy density.
However this requires costly cryogenic conditions during storage and transport. On the other
hand, storing methane in hydrates has approximately 3-4 times lower energy density than
LNG but the conditions are much milder. Methane hydrates, when palletized, can be stored in
a freezer at -20°C and atmospheric condition for about 2 weeks without significant hydrate
decomposition. Some gas hydrates show “self-preservation” properties, allowing storage well
below the equilibrium pressure at freezer temperatures, (Stern, 2001a; 2001b; Takeya et al.,
2001; Kuhs et al., 2004; Shimada et al., 2005), allowing storage and transport methane from
the gas field to the energy users. It has been suggested that the endothermic decomposition of
hydrate can be delayed when palletized and stored adiabatically as heat input is minimized.
The larger the hydrate pellet, the longer hydrate decomposition can be delayed (Takeya et al.,
2005).
The maximum possible and realizable methane content is obtained with sI hydrate but
this hydrate structure is the least stable of the methane-containing hydrates. Hence it is
essential to look at the overall performance in selecting the appropriate system that offers the
optimum in terms of gas content, stability condition, and the kinetics during formation and
decomposition. The addition of a second molecule to form sII and sH hydrate reduces the
equilibrium pressure. However the methane content is also reduced. The pressure reduction
depends on the type of guest molecule chosen and the concentration. For a double hydrate of
sII a much lower pressure is required however, the methane content decreases significantly,
too, which is unfavorable. In case of sH hydrate, the methane content reduction is not as large
with an appreciably reduction in pressure by a factor of two. Accordingly, there are many
recent efforts attempting to store methane in sH hydrate. Tuning gas hydrate properties to find
the compromise between storage capacity and pressure will be discussed at a later section.
There is additional gas storage potential in sH hydrate because the large cage might be
occupied by more than one methane molecules. Unfortunately this occurs only under
extremely high pressures and hence is not or practical interest. Currently the study of high
pressure methane hydrates of sH hydrate are relevant only to study the behavior of water and
gas molecules in the outer solar system (icy satellites/planetary studies). However further
studies are required to get the complete understanding on the feasibility of methane storage in
sH hydrate with multiple guest occupancy.
Hydrate Synthesis
The hydrate phase equilibria and gas storage capacity are constrained by the
thermodynamic conditions (temperature, pressure), crystallographic properties (structure,
cage size and shape) and guest molecule properties (size, shape). Since these properties are
often relatively well defined and fixed by specific external requirements, the optimization of
kinetics becomes an important topic from a practical viewpoint. It is indispensable to have
good recipes for hydrate synthesis. The rate of hydrate formation and the yield or conversion
level of hydrate are important from the economic point of view. The role of mass and heat
transfer is significant in order to ensure the progression of the reaction towards optimum
conversion. The contact between water and gas is also an important aspect for controlling the
hydrate conversion.
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 23
Hydrate formed from liquid water generally requires efficient mixing due to the poor
methane solubility in water. A rapid hydrate formation rate is essential for hydrate technology
to become useful to industry. Mixing can be achieved by using a mechanical mixer, a
magnetic bar coupled to a rotating magnet (Vysniauskas and Bishnoi, 1983; Lee et al.,
2005b), a static mixer (Tajima et al., 2004; 2005a; 2005b; 2006), a spray of water droplets
into a gas atmosphere (Tsuji et al., 2004; 2005), or bubbling the gas into the liquid water
phase (Ma et al., 2002; Takahashi et al., 2003). The performance of several contacting modes
is currently being investigated. Mori’s laboratory at Keio University (Japan) has proposed the
formation of hydrate by spraying water droplets into a gas atmosphere (Ohmura et al., 2002;
Fukumoto et al., 2001). From the hydrate formation rates measured for different systems it
was acknowledged that hydrate conversion was limited, especially after a certain amount of
solid crystals had been produced. The crystals have to be removed from the water circulation
line to prevent plugging the transmission lines. It is also challenging to avoid clogging the
nozzle. The other contacting mode suggested was the injection of micro-bubbles into bulk
water (Takahashi et al., 2003). The micro-bubble has a very long lifetime in the liquid which
allows sufficient time for hydrate formation. Static mixers have also been proposed recently
in a commercial hydrate plant to enhance the hydrate formation rate by increasing the contact
of gas and liquid (D’Aquino, 2003). Shell and tube heat exchangers are employed to remove
the heat generated by hydrate formation immediately. This technology was initially proposed
by Yamasaki’s group at AIST Tsukuba, Japan that claimed that much less energy
consumption is needed (Tajima et al., 2004; 2005a; 2005b; 2006). It has been reported that
the exposure of water on a cold metal surface may help to improve the kinetics (Matsuda et
al., 2006; Li et al., 2005) although hydrate conversion is limited when the surface becomes
covered by hydrates. All previously mentioned contacting modes work well mostly at the
early stages of hydrate formation. However complete water to hydrate conversion is unlikely
for most methods. There are always possibilities that the water molecules are occluded
between hydrate chunks and which limit the contact between gas and water, and hence the
extent of conversion. High power consumption is also required to create new interfaces
between water and gas and this increases the operational cost (except for the static mixer).
Hence additional stages may be required to convert un-reacted water or to separate the
hydrate from the slurry.
There are also several alternatives to hydrate synthesis that do not involve mixing. The
first one uses suitable surfactants to enhance the contact between water and gas and forms
porous hydrates (Zhong and Rogers, 2000; Sun et al., 2003; Link et al., 2003; Lin et al., 2004;
Wanatabe et al., 2005; Gayet et al. 2005). The addition of a small amount of sodium dodecyl
sulfate (SDS) anionic surfactant was reported to increase hydrate formation rates and water to
hydrate conversion by preventing the agglomeration of hydrate crystals (Sun et al., 2003;
Link et al., 2003; Lin et al., 2004; Wanatabe et al., 2005; Gayet et al. 2005). Hydrate growth
along the wall of the reaction vessel forms open pore structures that transport the water via
capillary action towards the hydrate crystallization front and this keeps the water-gas interface
free of hydrate. It was reported that 97% of water can be converted into hydrate (Link et al.,
2003; Gayet et al. 2005). SDS also enhances the decomposition kinetics (Lin et al., 2004).
Rapid hydrate formation towards full conversion was successfully reported by confining
water in porous media (silica gel) where ~90% of water could be converted into hydrate
within an hour when exposed to a carbon dioxide and hydrogen gas (Seo et al., 2005a; Lee et
al., 2005a). Unfortunately such kinetics study has never been reported for other systems
24 Peter Englezos, John Ripmeester and Robin Susilo
including methane hydrate. This contacting mode seems promising although the storage
density will be less because of the presence of silica gel. The most recently reported
contacting mode proposed is fluidization, giving better mass and heat transfer, but this has not
been tested experimentally (Servio et al., 2004). It seems that dispersing water in silica gels or
by adding SDS surfactant are the best approaches to synthesize hydrate from liquid water
considering the fast kinetics achieved, high water to hydrate conversion and lack of energy
requirements for mixing.
Despite the study on liquid water, hydrate can also be grown from ice. Early hydrate
formation kinetics study from ice was reportedly immeasurable due to limited growth (Hwang
et al., 1990). This is likely due to the formation of hydrate film that covers the ice surface and
exhibits significant mass transfer resistance. Thus a means of mixing is required to expose the
surface contact between ice and gas. Measurable rates were observed only when the ice was
melted (Hwang et al., 1990). This thermal ramping above the icepoint has been reported to
successfully produce pure methane hydrate in large quantities (Stern et al., 1998). Although
this is commonly employed in laboratory practice only, the thermal ramping procedure seems
attractive from the economical point of view to be applied in a larger scale because agitation
is not required. Kinetics studies on hydrate formation from ice have been reported recently
with in-situ measurements using diffraction and spectroscopy techniques (Susilo et al., 2006;
Kuhs et al., 2006; Staykova et al., 2003). Field emission-scanning electron microscopy (FE-
SEM) was also employed to examine the grain structure, and texture on ice/hydrate (Stern et
al., 2004). The formation of meso to macro porous on the surface of hydrate covered ice
particles was observed (Klapproth et al., 2003; Staykova et al., 2003).
Generally, hydrate growth from ice can be described into 2 stages (Susilo et al., 2006;
Kuhs et al., 2006; Staykova et al., 2003). The growth in the first stage is the formation of
hydrate film covering the ice surface which occurs relatively fast due to direct contact
between ice surface and the gas phase. This occurs until approximately 10% of ice is
converted into hydrate (Wang et al., 2002). Recently the presence of gas film was also
observed (Susilo et al., 2006). The gas concentration at the bulk and on the ice surface is not
equal. Mixing in the gas phase is required to remove the gas film resistance (Lee et al.,
2005c). The second growth stage can be divided into two steps. The first step is mass
transport from the bulk gas phase across the hydrate film, which is a slow process and
considered as the rate limiting step. The second step is the interfacial reaction at the hydrate-
ice interface. This hydrate reaction kinetics is the same as the well-known gas-solid reaction
with the product (ash) cover the solid interface which can be modeled with the shrinking core
model (Levenspiel, 1999). A more sophisticated model was developed and improved
(Salamantin et al., 1998; Staykova et al., 2003; Genov et al., 2004; Kuhs et al., 2006).
However the phenomena on temperature ramping cannot be described by the shrinking core
model. The melting of ice covered by the hydrate layer melts at the usual melting point
(Moudrakovski et al., 1999). Ice melting causes the pressure inside the hydrate shell to drop
as a result of different density between ice and water. Hence the hydrate on the inside shell
may decompose that leads to the breakage of the shell and more contact between unreacted
water and gas. The broken pieces from the hydrate shell may in turn act as the hydrate seeds
that further speed up the kinetics (Susilo et al., 2006).
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 25
Hydrate dissociates at temperature and pressure condition below the phase boundary.
Hence recovery of methane gas from hydrates can be achieved by increasing the temperature
or lowering the pressure. Hydrate decomposes faster at pressure and temperature conditions
further away from the three-phase line. The gas release rate is usually fast at the beginning
and decreases monotonically with time. However, methane hydrate is known to have
anomalies or also called as “self-preservation” effect at temperature between 240-268K (Stern
et al., 2001a; Stern et al., 2003; Kuhs et al., 2004). Surprisingly, this behavior is not observed
in methane-ethane in sII hydrate (Stern et al., 2003). The study on sH methane hydrates is not
available yet. This is rather interesting because apparently ice shielding for gas diffusion is
not the real explanation of the “self preservation” mechanism. Hence it remains a puzzle why
the decomposition of sI methane hydrate is delayed and others are not. Methane sI hydrate at
268K can be stored for more than two weeks at atmospheric condition and over a month at
pressure above 1MPa (Circone et al., 2004a). Rapid hydrate decomposition at temperature
above the icepoint is reported, that is governed by the intrinsic dissociation reaction (Sun and
Chen, 2006). This is advantageous because methane gas from the gas fields can be stored and
transported to energy importing countries in form of hydrate pellets. The hydrates can be
stored in a freezer at atmospheric condition for about 2 weeks or a slightly higher pressure
(10-20 atm) for longer storage time. A specialized-design hydrate carrier ship was proposed
(Ota et al., 2002).
It is also of practical interests to control hydrate decomposition. The early studies were
intended to understand the mechanism of hydrate decomposition when plugging in pipeline
was encountered. The hydrate crystals generally decompose by de-pressurization (Kelkar et
al., 1998, Peters et al., 2000; Hong et al., 2006). Thermal stimulation has also been considered
in order to provide strategies for methane recovery from the natural hydrate by thermal
stimulation (Selim and Sloan, 1989; Ji et al., 2001; Hong et al., 2003; Hong and Pooladi-
Darvish, 2005). Hydrate decomposition studies may also be applicable to gas storage in
hydrates.
The first hydrate decomposition kinetics model was reported from Bishnoi’s laboratory
that incorporates the intrinsic kinetics (Kim et al., 1987). Another rigorous model was also
reported by using Fourier’s law in one dimension with moving boundary problem due to
thermal conduction (Ullerich et al., 1987). The model assumes that the solid ice forms right
after hydrate decomposes that eventually melt into gas and water when heat is applied on the
wall. Hence there are two boundaries (ice-hydrate and ice-water) that move inward away
from the heat source. The extended model that incorporates the heat transfer effect and
intrinsic kinetics was also presented (Jamalludin et al., 1989). A model for hydrate
dissociation in porous media by thermal stimulation was presented (Selim and Sloan, 1989).
The model was improved over time by taking into account the porosity of the hydrate plug
and extended the dimension from Cartesian to cylindrical coordinates (Kelkar et al., 1998,
Peters et al., 2000). Most models predict the time required to completely dissociate the solid
phase and the decomposition rates at different pressure, temperature, or porosity. Simulation
results were generally in good agreement with the experimental data. The simulation
approach using molecular dynamics was also reported (English et al., 2005).The study on
hydrate decomposition in silica gel meso-porous (Aladko et al., 2006a) and mixed hydrate
with methane, ethane, and propane mixture (Kawamura et al., 2003; 2006) were also reported.
26 Peter Englezos, John Ripmeester and Robin Susilo
Methane in sI hydrate requires a certain pressure and temperature condition for long term
storage. Although the “self-preservation” effect does sound attractive, it may not be the
perfect solution because the hydrate is essentially in meta-stable condition. A more stable
hydrate is preferable especially from the technical, safety, and economical point of view.
Hence the challenge is to find an additive which can help stabilize the hydrate at lower
pressures and/or higher temperatures. A more stable hydrate can be achieved by introducing a
second guest molecule to form sII or sH hydrate. The obvious drawback is that the addition of
a second molecule implies a reduction in methane content in hydrates due to less accessible
cages for methane. The question is then, is it worthy to have a stable hydrate with a lesser
amount of methane content?
hydrate cages. Therefore there may be a change in hydrate composition or even the structure
as well over time. In a NMR kinetics study of methane-propane hydrate propane was found to
fully occupy the large cage with methane in the small cage only (Kini e al., 2004). The large
cages filling rate was twice faster than the small cages although the number of large cages is
only half of the small cages. Hence it was hypothesized that the methane occupancy was
small. A two-step hydrate formation was also reported from the pressure profile and gas
chromatography measurements (Uchida et al., 2004). It was found that sII hydrates was
formed in the first step but transform to sI hydrate in the second step. Kinetics study with
water spraying mode for methane (90%), ethane (7%), and propane (3%) mixture was also
reported (Tsuji et al., 2005a; 2005b). The gas uptake rates and the gas phase composition
were measured during the hydrate formation. It was found that the gas consumption rate for
methane in sI hydrate and gas mixture in sII hydrate are not noticeably different. The
composition of ethane and propane in the gas phase was found to decrease over time until it
reaches a steady state. Consequently the gas phase becomes methane-rich. These results
further suggest that methane content in sII hydrate is low hence storing methane in sII hydrate
seems not too promising. Tuning sII hydrate has been successfully demonstrated to increase
the methane content but at the expense of slower kinetics as well.
Methane in sH Hydrate
There are at least 26 large molecules known to form sH hydrate (Ripmeester and
Ratcliffe, 1990; Lee et al., 2006a). Hydrate phase equilibria at different temperature and
pressure conditions are widely available in literature for most sH hydrate formers with
methane (Mehta and Sloan, 1993; Hutz and Englezos, 1996; Sun et al., 2002; Ohmura et al.,
2003a; 2003c; 2005a; 2005c; Makino et al., 2004; Nakamura et al., 2003; Hara et al., 2005;
Ohmura et al., 2005a; 2005c). A hydrate phase equilibrium prediction method is also
available (Mehta and Sloan, 1996; Chen et al., 2003; Ma et al., 2005). Morphology studies of
sH hydrate systems were reported (Servio and Englezos, 2003b; Ohmura et al., 2005b). The
lattice constants of sH hydrate are dependent on the size of the LMGS (Takeya et al., 2006).
Single crystals of sH hydrate with adamantane (Kirchner et al., 2004) and methylcyclohexane
(Udachin et al., 2002) have been analyzed by X-Ray diffraction. The occupancy of the large
cage was determined to be full with the methane occupancy at around 80%. The solid-state
NMR analysis also confirms the occupancy values (Seo and Lee, 2003; Susilo et al., 2007a).
The actual gas content in sH hydrate was reported recently (Susilo et al., 2007a). It was found
that the methane content in sH hydrate is approximately 20% to 40% less than in sI hydrates
depending on the LMGS employed and pressure during the synthesis. This confirms the
prediction by using the thermodynamics model (Mooijer van Heuvel et al., 2000).
Because most sH hydrate former are poorly miscible in water, the presence of LMGS
may act as additional mass transfer barrier resulting a slow kinetics. Hence an efficient way to
maintain the contact among methane, water, and LMGS is important in synthesizing sH
hydrate. Mori’s laboratory suggested the water spraying technique (Ohmura et al., 2002; Tsuji
et al., 2004). Six LMGSs were tested: methylcyclohexane (MCH), 2,2-dimethylbutane
(neohexane/NH), tert-butyl methyl ether (TBME), 3-methyl-1-butanol (isoamyl alcohol), 3,3-
dimethyl-2-butanone (pinacolone), and 2-methylcyclohexanone (2MCH). Interestingly, the
rates of sH hydrate formation may exceed that of sI, especially for the system with TBME,
2MCH and pinacolone. However TBME is recommended due to economics (Tsuji et al.,
2004). It was found that the hydrate formation rates were dependent on the type of LMGS and
28 Peter Englezos, John Ripmeester and Robin Susilo
hence it was suggested that the formation rates are correlated with the solubility of LMGS in
water.
A systematic effort is underway in our laboratories on structure H hydrates. Three large
guest molecules (LMGS) were chosen: tert-butyl methyl ether (TBME), neohexane (NH), and
methylcyclohexane (MCH). Liquid-liquid equilibria (LLE) of water-LMGS and vapor-liquid-
liquid equilibria (VLLE) with methane were reported (Susilo et al., 2005). LMGS solubility
in water was found to increase with pressure but decrease with temperature as shown in
Figure 3. Methane solubility in water and LMGS also increase with pressure. This explains
why hydrates generally form at higher pressure and lower temperature. It is important to note
that TBME is two orders of magnitude more soluble in water than MCH or NH. Hence one
may expect hydrate to form faster with TBME as the LMGS followed by NH and MCH.
NH MCH TBME
40 6000
35
NH and MCH in water / ppm
5000
20 3000
15
2000
10
1000
5
0 0
0 500 1000 1500 2000
Pressure / kPa
NH MCH TBME
10 14000
9
NH and MCH in water / ppm
12000
8
TBME in water / ppm
7 10000
6 8000
5
4 6000
3 4000
2
2000
1
0 0
275.5 283.15 298.15
Temperature / K
Figure 3. Solubility data of neohexane (NH), methylcyclohexane (MCH), and tert-butyl methyl ether
(MTBE) in water. The upper figure is the vapor-liquid-liquid equilibria at 275.5K and the lower figure
is the liquid-liquid equilibria at atmospheric pressure. (Reprinted from Fluid Phase Equilibria (Susilo et
al., 2005), Copy right (2005) with permission from Elsevier).
A kinetic study in a well-stirred semi-batch reactor was conducted to determine the rate
of methane as shown in Figure 4 (Lee et al., 2005b). As seen, the system with TBME has the
shortest nucleation time and fastest hydrate growth rate followed by NH and MCH. This trend
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 29
agrees with that of the solubility measurements. The measured rates are also in good
agreement with the measurements by using the water spraying contacting mode (Tsuji et al.,
2004).
The rate of sI hydrate formation for comparison is also indicated. As seen, the nucleation
without LMGS is almost instantaneous however the rate is slower than TBME. The observed
rates are generally linear in the first 30 minutes however they tend to slow down as soon as
sufficient hydrate slurry is present in the suspension. This occurs when ~10% of water has
been converted into hydrate where the magnetic induced mixing cannot perform its job well.
Hence better contacting mode may be required that implies higher power consumption
required due to an increase in viscosity of the suspension. This is not encouraging because of
the low conversion.
0.20
Driving force =1.00 MPa at 275.50 K
CH4 moles consumed
0.15 CH4-H2O-TBME
O
-H 2
CH 4
O-MCH
CH 4-H 2
0.10
O-NH
CH 4-H 2
0.05
0.00
Time (min)
Figure 4. Methane uptake profile in sI and sH hydrate at 275.5K and 1MPa pressure above their
corresponding equilibrium pressure. (Reprinted from Energy and Fuels (Lee et al., 2005b), Copy right
(2005) with permission from American Chemical Society).
It is also indicated that there are two hydrate growth stages for the MCH system. By
employing NMR spectroscopy and imaging the methane partition and evolution in the gas,
liquid, and hydrate phase and ice was monitored over time. In addition, hydrate conversion,
methane diffusivity in LMGS and LMGS contact with ice were studied (Susilo et al., 2006).
It is noted at this point that inadequate mixing and high pressures requirements would be the
limitation with NMR. Hence the use of magic angle spinning (MAS) technique is impossible
which makes it difficult to resolve the spectra. It is also challenging to obtain acceptable
signal to noise ratio without acquiring the spectra longer than the actual kinetics. Hence
deuterated methane-d4 was used in our study to eliminate signal interference from the proton
of ice and LMGS. Besides, hydrate has to be synthesized from ice powder to eliminate the
long nucleation time of liquid water and better contact between LMGS and ice. Hence the
mass transfer resistance is due to the dissolution of methane in the liquid LMGS film
covering the ice powder. Methane in the dissolved liquid or gas phase and solid phase could
30 Peter Englezos, John Ripmeester and Robin Susilo
1.00
0.80
Intensity
0.60
0.40
Ice-CD4
0.20 Ice-TBME-CD4
Ice-NH-CD4
Ice-MCH-CD4
0.00
0 200 400 600 800 1000 1200
Time [min]
1.00
Ice-TBME-CD4
0.80 Ice-CD4
Ice-NH-CD4
Intensity
0.60 Ice-MCH-CD4
0.40
0.20
0.00
0 200 400 600 800 1000 1200
Time [min]
Figure 5. Normalized intensity evolution of methane in the gas + dissolved in LMGS liquid phase (left)
and hydrate phase (right) at 253K and ~4.5MPa. (Reprinted from J. Phys. Chem. B (Susilo et al., 2006),
Copy right (2006) with permission from American Chemical Society).
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 31
All intensities were calibrated so that the amount of methane in the hydrate phase can be
quantified. Hence the ice to hydrate conversion can be calculated assuming methane occupies
both small and medium cage by 80% (Udachin et al., 2002). The corresponding ice to hydrate
conversion is given in Figure 6. At 253K and ~4.5MPa, the maximum conversion achieved is
~35% for TBME that is still far from completion. The conversions for the systems with NH,
MCH and sI methane hydrate are even lower than 20%. Generally ice gets converted into
hydrate quickly during the first 5 hours before the rates slow down significantly. This is due
to the formation of hydrate film on ice surface that add resistances to the mass transfer.
Fortunately, the temperature ramping procedure that is commonly used in synthesizing sI
methane hydrate was also applicable in sH hydrate. Rapid hydrate formation was observed
when the temperature was ramped above the icepoint (Figure 6). The total conversion
obtained after 40 hours of reaction time was still far below completion unfortunately. This is
due to limited space in the NMR cell. The cell was fully loaded with ice particles to acquire
sufficient signal intensity and the formation of hydrate eventually requires volume expansion
due to methane inclusion. Thus there is not anymore space available for the unreacted ice to
get converted into hydrate.
60
Ice-TBME-CD4
50 Ice-NH-CD4
Hydrate Conversion [%]
Ice-CD4
Ice-MCH-CD4
40
30
20
10
0
0 300 600 900 1200 1500 1800 2100 2400
Time [min]
Figure 6. Ice to hydrate conversion rate, assuming both 512 and 435663 are 80% occupied with methane,
T = 253K for the first 20 hours and T = 274K afterward. (Reprinted from J. Phys. Chem. B (Susilo et
al., 2006), Copy right (2006) with permission from American Chemical Society).
Figure 7 shows the MCH contact with ice at different ice packing densities. Generally
LMGS may fill the free space between ice particles upon pressurization. However if the ice is
highly packed, The LMGS contact with the ice particles is poor. Thus some ice particles are
isolated and this limits hydrate conversion. The bulk measurement in an external pressure
vessel shows that complete hydrate conversion can be achieved with a thermal ramping
procedure when the ice is loosely packed (Susilo et al., 2007b). Surprisingly, a second hydrate
growth stage for the system with NH at 274K was observed. This is similar to the one
observed from the kinetic data mentioned earlier but for the MCH system. This is attributed to
rupture when the occluded water is transformed into hydrate. The rapid conversion from
32 Peter Englezos, John Ripmeester and Robin Susilo
melting ice into hydrate creates internal pressure due to volume expansion of hydrate
formation that eventually provides new access for the water to get converted into hydrate.
Figure 7. Proton density image of Ice+MCH with different degree of ice packing: the left images
correspond to highly packed ice and the right images correspond to not so highly packed ice. (Reprinted
from J. Phys. Chem. B (Susilo et al., 2006), Copy right (2006) with permission from American
Chemical Society).
The trend of hydrate formation rates changes between TBME and NH or MCH system
when the unreacted ice is melted (Susilo et al., 2007b). The TBME system shows a much
slower phase transformation towards full conversion at temperatures above the icepoint
whereas the other hydrophobic guests MCH and NH show a rapid conversion. It was
suggested that the high solubility of TBME in water lowers the water activity resulting the
slow kinetics. Hence the TBME acts as hydrate inhibitor the same way as other
thermodynamics inhibitor except TBME is also a hydrate guest molecule. The most recent
work from our group reported that methane occupancy in sH hydrate is dependent on the
LMGS type. The methane occupancy was higher for MCH and NH system than TBME when
synthesized at the same operating condition (Susilo et al., 2007a). However it is unsure if the
lower methane occupancy was due to differences in hydrate growth rates (kinetics) or
thermodynamics (stability). Further studies are required to elucidate this and it is currently
being investigated.
The contact between LMGS and ice was monitored through micro-imaging NMR
experiments. The polar guest (TBME) wets the ice surface whereas the hydrophobic guest
(NH and MCH) does not wet ice well as shown in Figure 8. The poor wetting between
hydrophobic molecule and ice limits the contact and hence the hydrate growth/conversion.
The fact that hydrate growth rates vary among different LMGS systems can be explained by
the difference observed from methane diffusivity and wetting between LMGS and ice. This
clarifies why the kinetics of the NH system is slower than in the TBME system although the
methane diffusivity in LMGS is similar. The system with MCH has the slowest kinetics
(nucleation time and hydrate growth rate) due to slow methane diffusivity and poor contact
between MCH and ice powder. It is clear that the contact between water and guest molecules
is very important in hydrate kinetics. The presence of LMGS does not necessarily limit the
mass transfer of methane from the bulk gas phase to the water interface because of high
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 33
Figure 8. Wetting and diffusion of LMGS (MCH and TBME) between ice particles at atmospheric
condition observed by 1H micro-imaging NMR. (Reprinted from J. Phys. Chem. B (Susilo et al., 2006),
Copy right (2006) with permission from American Chemical Society).
Multiple Occupancy in large cage. The challenge for methane storage in sH hydrate is to
increase the storage capacity by putting multiple methane molecules into the large cage. This
can be done at extremely high pressures although there is a disagreement about the number of
methane molecules in the large cage. The solid phase analysis obtained by X-Ray and neutron
diffraction suggested there are 5 methane molecules in the large cage (Loveday et al., 2003a)
but only 2-3 methane molecules reported from Raman spectroscopy (Kumazaki et al., 2004).
Molecular dynamics (MD) simulations are currently employed to verify the multiple guest
occupancy. Such comparison between MD simulation and experimental measurements was
presented for rare gases (Ne, Ar, Kr, Xe) in sH hydrate at high pressures (Alavi et al., 2006b;
Ogienko et al., 2006; Manakov et al., 2004). It still remains to be seen if sH hydrate tuning
can also be performed in the same way as sII hydrates. Recently it was reported that methane
can replace the methylcyclohexane in sH hydrate by pressurizing the hydrate with methane up
to 110 bar (Yeon et al., 2006).
Hydrate or gas to solids technology (GTS) is expected to be positioned between the LNG
and CNG technologies in terms of gas storage density, temperature and pressure condition.
The motivation though is the substantial reduction in the capital and operating cost. Hence
GTS technology may be utilized to recover the stranded gas and even compete with the
current established LNG and CNG technologies for larger gas fields. Japan leads the way in
the effort to commercialize GTS. Mitsui Engineering and Shipbuilding Corp., Inc (MES) has
completed a hydrate production pilot plant which is able to produce 600 kg of hydrate per
34 Peter Englezos, John Ripmeester and Robin Susilo
day. The most recent interest came from an Indian oil and gas company who has reached an
agreement with Aker Kvaerner (AK) for storing and transporting natural gas via hydrates. AK
is a Norwegian company who collaborates with Norwegian University of Science and
Technology (NTNU) and MES to develop, establish, and commercialize the GTS technology.
2. CO2 Capture from Treated Flue Gas (CO2/N2/O2) and Fuel Gas
(CO2/H2) Gas Mixtures
Global climate change and ocean acidification as a result of increased emissions of
carbon dioxide are now serious global environmental issues. Consequently, capture of carbon
dioxide from existing power plants along with improving energy efficiency in energy
utilization is now a necessary action plan to stabilize and reduce carbon emissions in the
atmosphere. There are two general approaches to CO2 capture. Post-combustion capture
refers to separating carbon dioxide from a flue gas (Klara and Srivastava, 2002). Pre-
combustion capture refers to capturing carbon prior to combustion. This would be a suitable
approach for new power plants in which the fossil fuel is gasified and then through the shift
reaction is converted into a stream of H2 and CO2 (Barchas and Davis, 1992). The CO2 can
then be removed for disposal. The resultant stream of H2 could be used in fuel cells and not
just in a gas turbine.
Processes for the removal of a component from a multi-component gaseous stream
include cryogenic fractionation, selective adsorption by solid adsorbents, gas absorption,
membrane separation etc. (Barchas and Davis, 1992; Kikkinides et al., 1993). It has been
estimated that, the cost of separation and disposal of CO2 from existing coal fired, air blown
boilers would increase the cost of electricity by about 75% (Hendriks and Turkenburg, 1990).
The cost of separation of CO2 alone reduces the power generation efficiency from 38 to 26%
(Chakma et al., 1995). Aaron and Tsouris (2005) have reviewed the processes for CO2
separation from flue gases. Liquid absorption using amines was considered the most
promising current method while some other methods are promising but too new for
comparison. The use of gas hydrates was considered on of the novel methods under
investigation for separating CO2 from flue gas (mixture of CO2, N2 and O2) or synthesis gas
(mixture of CO2 and H2).
Treated flue gas contains CO2, N2, O2. Since N2 and O2 form hydrate crystals at
approximately the same conditions the treated flue gas is considered a CO2/N2 mixture. Thus
in post-combustion capture from power plants the task is to separate CO2 from a CO2/N2
mixture in which the CO2 molar content is approximately 15-20 %. Pre-combustion capture
involves separation of CO2 from a mixture with H2 in which the CO2 molar concentration is
approximately 40 % (Klara and Srivastava, 2002; Kang and Lee, 2000; Englezos and Lee,
2005).
CO2 capture from flue gas mixtures. Figure 9 below shows a conceptual process based on
gas hydrate data from Linga et al. (2006; 2007a; 2007b). Following a one-stage hydrate
formation / decomposition process for the CO2/N2 mixture, a CO2-rich gas is obtained which
contains 57 % CO2 at 10 MPa. A second hydrate foration/decomposition stage is used with
the CO2-rich gas that results in a new CO2-rich stream containing about 83 % CO2. A hybrid
process is advocated where the lean CO2 streams are passing through a gas separation process
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 35
like a membrane one. The process is completed with a third gas hydrate process stage that
may result in 98-99 % CO2 stream.
N2
Membrane
17 % CO2
separation
Hydrate
83 % N2 CO2
Process
-stage 1
H2 O
Hydrate
Process
-stage 2 Hydrate
Process
-stage 3
H2 O 98-99 % CO2
for Disposal
H2 O
CO2 capture from fuel gas mixtures. Figure 10 shows a conceptual hydrate-based process
for the separation of CO2 from a fuel gas. Only two hydrate stages are needed in this case to
recover a 98-99 mol % of CO2 (Linga et al., 2007b; Kumar et al., 2006).
H2
Membrane
40 % CO2 separation
Hydrate
60 % H2 Process- CO2
stage 1
Hydrate
H2O Process-
stage 2
Figure 10. A hybrid hydrate-membrane process for CO2 recovery from fuel gas.
From an engineering standpoint the challenges are in some regards similar to those faced
in the natural gas storage area and discussed above. One has to find efficient ways of
contacting gas and water in order to scale up the process effectively. The above conceptual
36 Peter Englezos, John Ripmeester and Robin Susilo
processes are based on a laboratory stirred vessel operating in a semi-batch mode. Andersson
and Haines (2005) have used an industrial scale stirred vessel hydrate rector to produce
hydrates for storage in the ocean. An alternative approach is hydrate formation in a packed
bed with silica gel (Seo et al., 2005). Park et al. (2006b) reported phase equilibrium data
which led them to assume that a three-stage hydrate formation/decomposition process would
result a stream with more than 96 mol % CO2.
Finally, the use of additives is required to lower the operating conditions and reduce
compression costs. This should be done in a way that does not compromise the separation
factor. Use of tetrahydrofuran in the hydrate formation process has been suggested in the
literature in order to reduce the hydrate formation pressure for the CO2/N2 system (Kang and
Lee, 2000; Seo et al., 2005). A suitable additive is also required to lower the hydrate forming
conditions for the CO2/H2 system. One such additive might be propane (Kumar et al., 2006).
3. Hydrogen Storage
Hydrogen is a light gas that is easily combustible. It is a clean fuel with high energy
density per weight but low on volume. It is the preferred fuel for fuel-cell but hydrogen
storage is always the barrier for the development of fuel-cell technology (Ross, 2006; Chalk
and Miller, 2006). A safe way to store hydrogen in a reasonable energy density is always
challenging. Large energy density can be obtained when hydrogen gas is compressed or
liquefied. However both processes require extreme conditions that are not practical and safe.
Hydrate is one of the alternative technologies that are currently being explored for storing
hydrogen. This was realized since hydrogen was discovered to form sII hydrates back in 2002
(Mao et al., 2002). Two hydrogen molecules were reportedly able to fill the small cages and
four molecules in the large cages. Hence a total hydrogen content of 5 wt% is achievable in
hydrate. Hydrogen occupancies were also determined as a function of pressure and
temperature using neutron diffraction (Lokshin et al., 2004). It was found that the small cage
is singly occupied by hydrogen but the large cage can be occupied by two to four molecules
depending on the temperature. This was also supported by simulation work using the
molecular dynamics (Alavi et al., 2005) and lattice dynamics (Inerbaev et al., 2006). The
drawback is that pure hydrogen hydrates require high pressure condition that is not
practicable. Therefore addition of a second molecule that is able to reduce the formation
pressure significantly is needed.
Recently it was found that introducing tetrahydrofuran (THF) as a second guest molecule
may lower the required pressures significantly (Florusse et al., 2004). This is a real
breakthrough that increases the hopes for hydrogen storage via gas hydrates. Obviously lower
hydrogen content in hydrate is expected. A hydrate tuning strategy to increase the hydrate
content in hydrate was also reported (Lee et al., 2005a). Lowering the THF concentration
allows pressure reduction while maintaining reasonably high hydrogen content in hydrate up
to 4 wt%. Surprisingly this tuning procedure is not repeatable (Strobel et al., 2006; Hester et
al., 2006). It was reported that the maximum hydrogen content is about 1wt% only with all
THF occupy the large cage and only a single hydrogen molecule in the small cage at
moderate pressure (less than 60MPa). The molecular dynamics study confirms the stabilizing
effect of THF (Alavi et al., 2006a). The small cage is more favorable when there is only a
single hydrogen molecule inside. The unit cell volume increases significantly when a second
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 37
hydrogen molecule is added. Hence the hydrogen content is practically limited to ~1wt%. It is
unsure if the double occupancy of the small cage is observed as a meta-stable condition where
the hydrogen may escape or diffuse to another empty small cage or out from the hydrate cage.
Further studies are needed to clarify this discrepancy.
The requirement for hydrogen storage in USA is currently 6wt% although in certain
countries a lower storage density may be acceptable. It is well-known that the maximum
hydrogen storage capacity in sII hydrate is only 5wt%, which is a bit short from 6wt%. This
occurs only at a high pressure that is unrealistic for practical applications. A more realistic
pressure limits the hydrogen content around 1wt% only. Hence it is clear that storing
hydrogen in hydrate still have a long shot to go. The other options available are to employ
other sII guest molecules or to form sH hydrate. So far there is no publication on hydrogen
storage using other help guest than THF. The success finding of THF should open up the
possibility of other guest molecule to work the same way as THF or even better. Storing
hydrogen in sH hydrate is also attractive since the large cage of sH hydrate may allow much
more hydrogen molecule inside. However up to date there is no published work available yet
on this. Hence there are still many opportunities to explore the potential of hydrate for
hydrogen storage.
The other drawback for the hydrate tuning that is worthwhile to mention is the slow
reaction time. Significant improvement in the kinetics was accomplished by dispersing the
water in a confined space using silica beads (Lee et al., 2005a). Full hydrate conversion can
be achieved within an hour with the proposed methodology where the normal reaction time
takes a week or more. However more studies are required to verify this.
using water droplets or water contained in cylindrical glass columns. It was found that the
water droplet containing the strong commercial inhibitor was found to collapse prior to
nucleation and spread out on a Teflon surface. Subsequently, hydrate was formed as a layer
on the surface. Catastrophic growth and spreading of the hydrate crystals was also observed
during hydrate formation in the glass columns in the presence of the kinetic inhibitor. Finally,
when polyethylene oxide (PEO) was added into the kinetic inhibitor solution the memory
effect on the induction time decreased dramatically (“suppression of memory”). Antifreeze
protein from winter flounder was also found to impact the memory effect. In fact Zeng et al.
(2006a) reported that the memory effect was eliminated.
The knowledge of the mechanism of action is required in order to synthesize more
effective inhibitors. Effectiveness refers to the ability to act increased degrees of
undercooling. Environmental considerations have also motivated the search for biologically
derived inhibitors instead of synthetic ones (Kelland, 2006; Zeng et al. 2006b)
8. Other Applications.
The fact that solutes are excluded from the hydrate lattice motivated work on using
hydrates for the desalination of seawater and the concentration of aqueous solutions in
general. Most of that work occurred in the 1960s and 70s and was reviewed by Englezos
(1993). More recent work dealt with water recovery from pulp mill effluents (Ngan and
Englezos, 1996); and concentration of liquid foods (Purwanto et al. 2001). The motivation for
such work is potential energy savings compared with evaporation and in the case of food
applications the preservation of the aroma volatile compounds. Other idea involving hydrates
is the photo-catalysis of methane hydrate (Taylor, 2005) and as a medium for cool energy
storage. This is particularly attractive to store the cool energy during the off-peak period and
discharge it when needed. The hydrate process can be implemented in a closed-loop vapor-
compression refrigerator where the hydrate crystallizer may replace a conventional evaporator
(Mori and Mori, 1989). The refrigerant in the evaporator can be exposed to water where the
energy released by the evaporation of the refrigerant is absorbed by the water to form hydrate.
Hence the refrigerant which is a volatile organic compound may be stored in hydrate while it
40 Peter Englezos, John Ripmeester and Robin Susilo
is not in use. The hydrate can be decomposed to cool off the air due to the endothermic
reaction of hydrate decomposition.
Concluding Remarks
Clathrate of gas hydrates are a research topic of interest to a variety of disciplines in
addition to chemical engineering. In most cases the issues are complex and require the
collaborative effort of scientists and engineers. In this work we have described the various
hydrate topics of interest to chemical engineering research. These areas include the
development of clean energy technologies such as natural gas transport and storage, capture
of carbon dioxide from flue gases, capture of carbon dioxide from fuel gases and recovery of
hydrogen, hydrogen storage. Natural gas storage and transport have reached pilot plant stage
in Japan while the CO2 capture technologies and hydrogen storage are at the laboratory stage.
The recovery of energy from hydrates in the earth and the relationship of hydrates in the earth
with global climate change are briefly mentioned since they are covered extensively
elsewhere. Flow assurance in hydrocarbon pipelines is currently the only area where hydrates
are involved in an industrial scale. The shift there is towards finding kinetic inhibitors derived
from biological sources rather than using synthetic macromolecules. The motivation is
environmental considerations. Other applications such as aqueous solution concentration and
cool energy storage are also mentioned. The emphasis is on exposing the reader to the state of
the art on the various topics and highlighting research needs.
References
[1] Aaron, D. & Tsouris, C. (2005). Separation of CO2 from Flue Gas: A Review.
Separation Science and Technology, 40 (1-3), 321-348.
[2] Abdalla, B.K. & Abdullatef, N.A. (2005). Simulation and economic evaluation of
natural gas hydrates [NGH] as an alternative to liquefied natural gas [LNG]. Catalysis
today, 106, 256-258.
[3] Aladko, E.Y. Dyadin, A.Y. Fenelonov, V.B. Larionov, E.G. Manakov, A.Y. Mel’gunov,
M.S. & Zhurko, F.V. (2006a). Formation and decomposition of ethane, propane, and
carbon dioxide hydrates in silica gel mesopores under high pressure. J. Phys. Chem.
B, 110 (39), 19717-19725.
[4] Aladko, E.Y. Ancharov, A.I. Goryainov, S.V. Kurnosov, A.V. Larionov, E.G.
Likhacheva, A.Y. Manakov, A.Y. Potemkin, V.A. Sheromov, M.A. Teplykh, A.E.
Voronin, V.I. & Zhurko, F.V. (2006b). New Type of Phase Transformation in Gas
Hydrate Forming System at High Pressures. J. Phys. Chem. B, 110 (42), 21371-
21376.
[5] Alavi, S. Ripmeester, J.A. & Klug, D.D. (2005). Molecular-dynamics study of structure
II hydrogen clathrates. J. Chem. Phys., 123, 024507.
[6] Alavi, S. Ripmeester, J.A. & Klug, D.D. (2006a). Molecular-dynamics simulations of
binary structure II hydrogen and THF clathrates. J. Chem. Phys., 124 (1), 014704.
[7] Alavi, S. Ripmeester, J.A. & Klug, D.D. (2006b). Stability of rare gas structure H
clathrate hydrates. J. Chem. Phys., 125 (10), 104501.
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 41
[26] Chakma, A. Mehrotra, A.K. & Nielsen, B. (1995). Comparison of chemical solvents for
mitigating CO2 emissions from coal-fired power plants. Heat Recovery Systems &
Chp., 15 (2), 231-240.
[27] Chalk, S.G. & Miller, J.F. (2006). Key challenges and recent progress in batteries, fuel
cells, and hydrogen storage for clean energy systems. J. Power Resources, 159 (1),
73-80.
[28] Chatti, I. Delahaye, A. Fournaison, L. & Petitet, J.P. (2005). Benefits and Drawbacks of
Clathrate Hydrates: a Review of their areas of interest. Energy Conversion and
Management, 46, 1333-1343.
[29] Chazallon, B. & Kuhs, W. F. (2002). In situ structural properties of N-2-, O-2-, and air-
clathrates by neutron diffraction. J. Chem. Phys., 117 (1), 308-320.
[30] Chen, G.-J. Sun, C.-Y., Guo, T.-M. (2003). Modelling of the formation conditions of
structure-H hydrates. Fluid Phase Equilibria, 204 (1), 107-117.
[31] Chen, A.B. Zhang, W.P. Lan, X.J. Zheng, H. Liu, M.X. Han, X.W. Bao, X.H. (2006). In
situ high-pressure NMR observations on the formation and dissociation of methane
hydrate. Chinese Chemical Letters, 17 (6), 791-794.
[32] Chou, I.-M. Sharma, A. Burrus, R.C. Shu, J. Mao, H. Hemley, R.J. Goncharov, A.F.
Stern, L.A. Kirby, S.H. (2000). Transformation in Methane Hydrate. PNAS, 97 (25),
13484-13487.
[33] Circone, S. Stern, L.A. & Kirby, S.H. (2004a). The effect of elevated methane pressure
on methane hydrate dissociation. American Mineralogist, 89, 1192–1201.
[34] Circone, S. Kirby, S.H. & Stern, L.A. (2005). Direct measurement of methane hydrate
composition along the hydrate equilibrium boundary. J. Phys. Chem. B, 109, 9468-
9475.
[35] Clarke, M.A. & Bishnoi, P.R. (2004). Determination of the intrinsic rate constant and
activation energy of CO2 gas hydrate decomposition using in-situ particle size
analysis. Chem. Eng. Sci., 59 (14), 2983-2993.
[36] Clarke, M.A. & Bishnoi, P.R. (2005). Determination of the intrinsic kinetics of CO2 gas
hydrate formation using in situ particle size analysis. Chem. Eng. Sci., 60 (3), 695-
709.
[37] Collett, T.S. (2005). Results at Mallik highlight progress in gas hydrate energy resource
research and development. Petrophysics, 46 (3), 237-243.
[38] Collins, M.J. Ratcliffe, C.I. Ripmeester, J.A. (1990). Nuclear magnetic resonance
studies of guest species in clathrate hydrates: line-shape anisotropies, chemical shifts,
and the determination of cage occupancy ratios and hydration numbers. J. Phys.
Chem., 94 (1), 157-162.
[39] D’Aquino, R. (2003). A wake-up call in water management. CEP magazine 99 (3),
10-17.
[40] Dallimore, S.R. Collett, T.S. (Eds.). (2005) Scientific Results from the Mallik 2002 Gas
Hydrate Production Research Well Program, Mackenzie Delta, Northwest
Territories, Canada, Ottawa, Canada, Geological Survey of Canada Bulletin 585,
including CD.
[41] Dalmazzone, D. Hamed, N. Dalmazzone, C. & Rousseau, S. (2006). Application of high
pressure DSC to the kinetics of formation of methane hydrates in water-in-oil
emulsion. J. Thermal Analysis & Calorimetry, 85 (2), 361-368.
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 43
[42] Davidson, D.W. Gas Hydrates. In: Frank, F. Eds. Water: A Comprehensive Treatise.
New York, Plenum Press, 1973, Vol. 2, Chapter 3, 115.
[43] Davy, H. (1811). The Bakerian Lecture. On Some of the Combinations of Oxymuriatic
Gas and Oxygene, and on the chemical Relations of these Principle, to inflammable
Bodies. Phil. Trans. Roy. Soc. London, 101, Part I, 1-35.
[44] de Deugd, R.M. Jager, M.D. & de Swaan Arons, J. (2001). Mixed Hydrates of Methane
and Water-Soluble Hydrocarbons Modeling of Empirical Results. AIChE J., 47(3),
693-704.
[45] Deaton, W.M. Frost, E.M. Jr. (1946). Gas hydrates and their relation to the operation of
natural-gas pipe lines. U.S. Bureau of Mines Monograph, 8, 101.
[46] Delsemme, A.H., Swings, P. (1952) Gas hydrates in cometary nuclei and interstellar
grains, Ann. Astrophys, 15, 1.
[47] Delsemme, A.H., Wenger, A. (1970) Physico-chemical phenomena in comets-I:
Experimental study of snows in a cometary environment, Planet Space. Sci., 18,
709-715.
[48] Dholabhai, P.D. Kalogerakis, N. & Bishnoi, P.R. (1993). Kinetics of methane hydrate
formation in aqueous-electrolyte solutions. Can. J. Chem. Eng., 71 (1), 68-74.
[49] Dholabhai, P.D. Kalogerakis, N. & Bishnoi PR. (1992). Inhibition of gas hydrate
deposition in offshore pipelines: effect of methanol. Preprint of Paper # CIM 92-83,
Annual Meeting of The Petroleum Society of CIM, Calgary, Alberta, June 7-10.
[50] Dickens, G. R. Paull, C. K. Wallace, P. & the ODP Leg 164 Scientific Party. (1997).
Direct measurement of in situ methane quantities in a large gas-hydrate reservoir.
Nature, 385, 426.
[51] Dickens, G. R. O’Neil, J. R. Rea, D. K. & Owen, R. M. (1995). Dissociation of oceanic
methane hydrate as a cause for the carbon isotope excursion at the end of the
Paleocene. Paleoceanography, 10, 965.
[52] Englezos, P. Kalogerakis, N. Dholabhai, P.D. & Bishnoi, P.R. (1987a). Kinetics of
formation of methane and ethane gas hydrates. Chem. Eng. Sci., 42 (11), 2647-2658.
[53] Englezos, P. Kalogerakis, N. Dholabhai, P.D. & Bishnoi, P.R. (1987b). Kinetics of gas
hydrate formation from mixtures of methane and ethane gas. Chem. Eng. Sci,. 42 (11),
2659-2666.
[54] Englezos, P. Huang, Z. & Bishnoi, P.R. (1991). Prediction of Natural Gas Hydrate
Formation Conditions in the Presence of Methanol Using the Trebble-Bishnoi
Equation of State. The Journal of Canadian Petroleum Technology, 30, 148-155.
[55] Englezos, P. (1993). Clathrate Hydrates, Ind. Eng. Chem. Res., 32, 1251-1274.
[56] Englezos, P. (1996). Nucleation and growth of gas hydrate crystals in relation to ''kinetic
inhibition''. Revue Del Institut Francais Du Petrole, 51 (6), 789-795.
[57] Englezos, P. & Lee, J. D. (2005). Gas Hydrates: A Cleaner Source of Energy and
Opportunity for Innovative Technologies. Korean J. Chem. Eng., 22, 671-681.
[58] English, N.J. Johnson, J.K. & Taylor, C.E. (2005). Molecular-dynamics simulations of
methane hydrate dissociation. J. Chem. Phys., 123 (24), Art. No. 244503.
[59] Florusse, L. J. Peters, C. J. Schoonman, J. Hester, K. C. Koh, C. A. Dec, S. F. Marsh,
K.N. & Sloan, E.D. (2004). Stable Low-Pressure Hydrogen Clusters Stored in a
Binary Clathrate Hydrate. Science, 306, 469.
[60] Fu, B. Neff, S. Mathur, A. & Bakeev, K. (2002). Application of low-dosage hydrate
inhibitors in deepwater operations. SPE Production & Facilities, 17(3), 133-137.
44 Peter Englezos, John Ripmeester and Robin Susilo
[61] Fukumoto, K. Tobe, J. Ohmura, R. & Mori, (2002). Y.H. Hydrate formation using water
spraying in a hydrophobic gas: A preliminary study. AIChE J., 47, 1899-1904.
[62] Gao, S. House, W. & Chapman, W.G. (2005). NMR/MRI Study of Clathrate Hydrate
Mechanisms. J. Phys. Chem. B, 109, 19090-19093.
[63] Gayet, P. Dicharry, C. Marion, G. Graciaa, A. Lachaise, J. & Nesterov, A. (2005).
Experimental determination of methane hydrate dissociation curve up to 55 MPa by
using a small amount of surfactant as hydrate promoter. Chem. Eng. Sci., 60 (21),
5751-5758.
[64] Genov, G. Kuhs, W.F. Staykova, D.K. Goreshnik, E. & Salamatin, A.N. (2004).
Experimental studies on the formation of porous gas hydrates. American
Mineralogist, 89, 1228-1239.
[65] Glasby, G.P. (2003). Potential impact on climate of the exploitation of methane
hydrate deposits offshore. Marine &Petroleum Geology 20 (2), 163-175.
[66] Gudmundsson, J.S., Parlaktuna, M., Khokhar, A.A. (1994). Storing Natural Gas as
Frozen-Hydrate. SPE Prod. Facilities 9 (1), 69-73.
[67] Gudmundsson, J.S., Andersson, V., Levik, O.I., Mork, M. (2000). Hydrate technology
for capturing stranded gas. Ann. N.Y. Acad. Sci., 912, 403-410.
[68] Hammerschmidt, E.G. (1934). Formation of Gas Hydrates in Natural Gas
Transmission Lines. Ind. Eng. Chem., 26, 851-855.
[69] Hand, K.P. Chyba, C.F. Carlson, R.W. & Cooper, J.F. (2006). Clathrate Hydrates of
Oxidants in the Ice Shell of Europa. Astrobiology, 6(3), 463-482.
[70] Handa, Y.P. (1985). Heat capacities in the range 95 to 260 K and enthalpies of fusion
for structure-II clathrate hydrates of some cyclic ethers. J. Chem. Thermodynamics, 17
(3), 201-208.
[71] Handa, Y.P. (1986a). Compositions, enthalpies of dissociation, and heat capacities in
the range 85 to 270 K for clathrate hydrates of methane, ethane, and propane, and
enthalpy of dissociation of isobutane hydrate, as determined by a heat-flow
calorimeter. J. Chem. Thermodynamics, 18 (10), 915-921.
[72] Handa, Y.P. (1986b). Calorimetric determinations of the compositions, enthalpies of
dissociation, and heat capacities in the range 85 to 270 K for clathrate hydrates of
xenon and krypton. J. Chem. Thermodynamics, 18 (9), 891-902.
[73] Hara, T. Hashimoto, S. Sugahara, T. & Ohgaki, K., (2005). Large pressure depression
of methane hydrate by adding 1,1-dimethylcyclohexane. Chem. Eng. Sci., 60, 3117-
3119.
[74] Hatzikiriakos, S. G. & Englezos, P. (1993). The relationship between global warming
and methane gas hydrates in the earth. Chem. Eng. Sci., 48 (23), 3963-3969.
[75] Hendriks, C.B. K. & Turkenburg, W. (1990). Climate and Energy. in, (Dordrec, The
Netherlands), Kluwer Academic Press, 125-188.
[76] Hester, K.C. & Sloan, E.D. (2005). sII Structural Transitions from Binary Mixtures of
Simple sI Formers. Int. J. Thermophysics, 26 (1), 95-106.
[77] Hester, K.C. Strobel, T.A. Sloan, E.D. Koh, C.A. Huq, A. & Schultz, A.J. (2006).
Molecular hydrogen occupancy in binary THF-H-2 clathrate hydrates by high
resolution neutron diffraction. J. Phys. Chem., B 110 (29), 14024-14027.
[78] Hong, H. Pooladi-Darvish, M. & Bishnoi, P.R. (2003). Analytical modelling of gas
production from hydrates in porous media. J. Canadian petroleum technology, 42
(11), 45-56.
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 45
[97] Kang, S.P. & Lee, H. (2000). Recovery of CO2 from flue gas using gas hydrate:
Thermodynamic verification through phase equilibrium measurements. Environ. Sci.
Technol., 34(20), 4397-4400.
[98] Kang, S.-P. & Lee, H. (2001). Enthalpies of dissociation of clathrate hydrates of
carbon dioxide, nitrogen, (carbon dioxide + nitrogen), and (carbon dioxide + nitrogen
+ tetrahydrofuran). J. Chem. Thermodynamics, 33 (5), 513-521.
[99] Katz, M. E. Pak, D. K. Dickens, G. R. & Miller, K. G. (1999). The source and fate of
massive carbon input during the latest paleocene thermal maximum. Science, 286,
1531.
[100] Kawamura, T. Ohga, K. Higuchi, K. Yoon, J.H. Yamamoto, Y. Komai, T. & Haneda,
H. (2003). Dissociation Behavior of Pellet-Shaped Methane-Ethane Mixed Gas
Hydrate Samples. Energy and Fuels, 17, 614-618.
[101] Kawamura, T. Sakamoto, Y. Ohtake, M. Yamamoto, Y. Komai, T. Haneda, H. Yoon,
J.-H. & Ohga, K. (2006). Dissociation behavior of pellet shaped mixed gas hydrate
samples that contain propane as a guest. Energy Convers. Manage., 47 (15-16), 2491-
2498.
[102] Kelkar, S.K. Selim, M.S. & Sloan, E.D. (1998). Hydrate dissociation rates in
pipelines. Fluid Phase Equilibria, 150, 371-382.
[103] Kelland, M.A. (2006). History of the Development of Low Dosage Hydrate
Inhibitors. Energy and Fuels, 20(3), 825-847.
[104] Kerr, R. (1997). Did a blast of sea-floor gas hydrates usher in a new age? Nature,
275,1267.
[105] Khokkar, A.A. Gudmundsson, J.S. & Sloan, E.D. (1998). Gas storage in structure H
hydrates. Fluid Phase Equilibria, 150-151, 383-392.
[106] Kikkinides, E.S. Yang, R.T. & Cho, S.H. (1993). Concentration and recovery of
carbon dioxide from flue gas by pressure swing adsorption. Industrial & Engineering
Chemistry Research, 32 (11), 2714-2720.
[107] Kim, H.C. Bishnoi, P.R. Heidemann, R.A. & Rizvi, S.S.H. (1987). Kinetics of
methane hydrate decomposition. Chem. Eng. Sci., 42, 1645-1653.
[108] Kim, D.-Y. Lee, J.W. Seo, Y.-T. Ripmeester, J.A. & Lee. H. (2005). Structural
Transition and Tuning of tert-Butylamine Hydrate. Angew. Chem. Int. Ed., 44 (47),
7749-7752.
[109] Kim, D.-Y. Park, J. Lee, J.-W. Ripmeester, J.A. & Lee, H. (2006). Critical guest
concentration and complete tuning pattern appearing in the binary clathrate hydrates.
J. Am. Chem. Soc., 128 (48), 15360-15361.
[110] Kini, R.A. Dec, S.F. & Sloan, E.D. Jr. (2004). Methane + Propane Structure II
Hydrate Formation Kinetics. J. Phys. Chem. A, 108, 9550-9556.
[111] Kirchner, M.T. Boese, R. Billups, W.E. & Norman, L.R. (2004). Gas Hydrate Single-
Crystal Structure Analyses. J. Am. Chem. Soc., 126, 9407-9412.
[112] Klapproth, A. Goreshnik, E. Staykova, D. Klein, H. & Kuhs, W.F. (2003). Structural
studies of gas hydrates. Can. J. Phys., 81, 53-518.
[113] Klara, S.M. & Srivastava, R.D. (2002). U.S. DOE Integrated Collaborative
Technology Development Program for CO2 Separation and Capture. Environmental
Progress, 21(4), 247-253.
[114] Klauda, J., B. & Sandler, S.I. (2005). Global distribution of methane hydrate in ocean
sediment. Energy & Fuels, 19, 459-470.
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 47
[132] Lee, H. Lee, J. Kim, D.Y. Park, J. Seo, Y.-T. Zeng, H. Moudrakovski, I.L. Ratcliffe,
C.I. & Ripmeester, J.A. (2005a). Tuning clathrate hydrates for hydrogen storage.
Nature, 434, 743-746.
[133] Lee, J.-D. Susilo, R. & Englezos, P. (2005b). Kinetics of Structure H Gas Hydrate.
Energy and Fuels, 19, 1008-1015.
[134] Lee, J.D. Susilo, R. Englezos, P. (2005c). Methane-ethane and methane-propane
hydrate formation and decomposition on water droplets. Chem. Eng. Sci., 60, 4203-
4212.
[135] Lee, J.D. & Englezos, P. (2005) Enhancement of the Performance of Gas Hydrate
Kinetic Inhibitors with Polyethylene Oxide. Chem Eng Sci, 60, 5323-5330.
[136] Lee, J.D. & Englezos, P. (2006). Unusual kinetic inhibitor effects on gas hydrate
formation. Chem. Eng. Sci. 61 (5), 1368-1376.
[137] Lee, J. Lu, H. Moudrakovski, I.L. Ratcliffe, C.I. & Ripmeester, J.A. (2006a). n-
pentane and n-hexane as coguests in structure-H hydrates in mixtures of 2,2-
dimethylbutane and methane. Angew. Chem. Int. Ed., 45 (15), 2456-2459.
[138] Lee, J.D. Song, M. Susilo, R. & Englezos, P. (2006b). Dynamics of Methane-Propane
Clathrate Hydrate Crystal Growth from Liquid Water with or without the Presence of
n-Heptane. Crystal Growth & Design, 6 (6), 1428-1439.
[139] Lekvam, K. & Ruoff, P. (1993). A reaction kinetic mechanism for methane hydrate
formation in liquid water. J. Am. Chem. Soc., 115 (19), 8565-8569.
[140] Lekvam, K. & Bishnoi, P.R. (1997). Dissolution of methane in water at low
temperatures and intermediate pressures. Fluid Phase Equilibria, 131, 297-309.
[141] Levenspiel, O. (1999). Chemical Reaction Engineering, 3rd ed., New York, NY:
Wiley & Sons.
[142] Li, J. Liang, D. Guo, K. & Wang, R. (2005). The influence of additives and metal
rods on the nucleation and growth of gas hydrates. J. Coll. & Int. Sci., 283 (1), 223-
230.
[143] Li, X.-S. & Englezos, P. (2006). Prediction of Gas Hydrate Formation Conditions in
the Presence of Methanol, Glycerol, Ethylene glycol and Triethylene glycol with the
SAFT Equation of State. Ind Eng Chem Res., 45 (6), 2231-2137.
[144] Li, X.-S. Wu, H.-J. Li, L.-G. Feng, Z.-P. Tang, L.-G. & Fan, S.-H. (2007). Hydrate
dissociation conditions for gas mixtures containing carbon dioxide, hydrogen,
hydrogen sulfide, nitrogen, and hydrocarbons using SAFT. J Chem Thermodynamics.
39, 419-425.
[145] Lin, W., Chen, G.-J., Sun, C.-Y., Guo, X.-Q., Wu, Z.-K., Liang, M.-Y., Chen, L.-T.,
Yang, L.-Y., Effect of surfactant on the formation and dissociation kinetic behavior of
methane hydrate, Chem. Eng. Sci. 59: 4449-4455 (2004).
[146] Linga, P. Kumar, R. & Englezos, P. (2006). Capture of Carbon Dioxide from
Conventional Power Plants or from Integrated Gasification Plants through Gas
Hydrate Formation/Dissociation. Journal of Energy & Climate Change, 1 (2), 75-82.
[147] Linga, P. Kumar, R. & Englezos P. (2007a). Gas Hydrate formation from
Hydrogen/carbon dioxide and nitrogen/carbon dioxide gas mixtures. accepted for
publication in Chem. Eng. Sci., April 26, 2007
[148] Linga, P. Kumar, R. & Englezos P. (2007b). The clathrate hydrate process for post
and pre combustion capture of carbon dioxide. J Hazardous Materials. revised
submitted May 10, 2007
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 49
[165] Makogon, T.Y. & Sloan, E.D. Jr. (1994). Phase Equilibrium for Methane Hydrate
from 190 to 262 K. J. Chem. Eng. Data, 39 (2), 351-353.
[166] Makogon, T.Y. Mehta, A.P. & Sloan, E.D. Jr. (1996). Structure H and Structure I
Hydrate Equilibrium Data for 2,2-Dimethylbutane with Methane and Xenon. J. Chem.
Eng. Data, 41, 315-318.
[167] Manakov, A.Y. Voronin, V.I. Kurnosov, A.V. Teplykh, A.E. Komarov, A.Y. &
Dyadin, A.Y. (2004). Structural Investigations of Argon Hydrates at Pressures up to
10 kbar. J. Inclusion Phenom. Mol. Recognit. Chem., 48 (1-2), 11-18.
[168] Mao, W.L. Mao, H. Goncharov, A.F. Struzhkin, V.V. Guo, Q. Hu, J. Shu, J. Hemley,
R.J. Somayazulu, M. & Zhao, Y. (2002). Hydrogen clusters in clathrate hydrate.
Science, 297 (5590), 2247-2249.
[169] Marshall, C.B. Fletcher, G.L. & Davies, P.L. (2004). Hyperactive antifreeze protein in
a fish. Nature, 429, 153.
[170] Matsuda, S. Tsuda, H. & Mori, Y.H. (2006). Hydrate Formation Using Water
Spraying onto a Cooled Solid Surface in a Guest Gas. AIChE J., 52 (8), 2978-2987.
[171] Max, M.D. Johnson, A. H. Dillon, W. P. (2006). Economic Geology of Natural Gas
Hydrates. The Netherlands: Springer-Verlag.
[172] McMullan, R.K. & Jeffrey, G.A. (1965). Polyhedral Clathrate Hydrates. IX. Structure
of Ethylene Oxide Hydrate. J. Chem. Phys., 43 (8), 2725-2732.
[173] Mehta, A.P. & Sloan, E.D. Jr. (1993). Structure H Hydrate Phase Equilibria of
Methane + Liquid Hydrocarbon Mixtures. J. Chem. Eng. Data, 38, 580-582.
[174] Miller, B. & Strong, E.R. (1946). Am. Gas Assn. Monthly, 29(2), 63.
[175] Mochizuki, T. (2003). Clathrate hydrate formation at liquid/liquid interface under
shear water flow. J. Crystal Growth, 249, 372–380.
[176] Mochizuki, T. & Mori, Y.H. (2006). Clathrate-hydrate film growth along
water/hydrate-former phase boundaries - numerical heat-transfer study. J. Crystal
Growth, 290 (2), 642-652.
[177] Mooijer van Heuvel, M.M. Peters. C.J. & de Swaan Arons, J. (2000). Influence of
water-insoluble organic components on the gas hydrate equilibrium conditions of
methane. Fluid Phase Equilibria, 172, 73-91.
[178] Mori, T. & Mori, Y.H. (1989a). Characterization of gas hydrate formation in direct-
contact cool storage process. Int. J. Refrigeration, 12 (5), 259-265.
[179] Mori, T. & Mori, Y.H. (1989b). Formation of gas hydrate with CFC alternative R-
134a, AIChE J., 35 (7), 1227-1228.
[180] Mori, Y.H. (2001). Estimating the thickness of hydrate films from their lateral growth
rates: application of a simplified heat transfer model. J. Crystal Growth, 223, 206-
212.
[181] Mori, Y.H. (2003). Recent Advances in Hydrate-based Technologies for Natural gas
Storage-A Review. J. Chem. Ind. Eng. (China), 54, Suppl., 1−17.
[182] Moridis, G. J. (2003). Numerical studies of gas production from methane hydrates.
paper SPE 75691, SPE Journal, 8(4), 359.
[183] Moridis, G. J. (2004). Numerical studies of gas production from class 2 and class 3
hydrate accumulations at the Mallik site, Mackenzie Delta, Canada. SPE Reservoir
Evaluation & Engineering, 7(3), 175.
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 51
[215] Pooladi-Darvish, M. (2004). Gas Production from Hydrate Reservoirs and its
Modelling. Distinguished author series, Journal of Petroleum Technology (JPT), 65.
[216] Purwanto, Y.A. Oshita, S. Seo Y. & Kawagoe, Y. (2001). Concentration of Liquid
Foods by the Use of Gas Hydrate. J. of Food Engineering, 47, 133-138.
[217] Rautenbach, R. & Pennings, P. (1973). Development and optimization of a hydrate
process for desalination of seawater. Chemie Ingenieur Technik, 45 (5), 259-264.
[218] Reeburgh, W.S. (2003). Global Methane Biogeochemistry, Treatise on Geochemistry.
Vol 4, Editor: Ralph F. Keeling. Executive Editors: Heinrich D. Holland and Karl K.
Turekian. pp. 347. ISBN 0-08-043751-6. Elsevier, 65-89.
[219] Ripmeester, J.A. & Davidson, D.W. (1977). Some new clathrate hydrates. Mol. Crys.
Liq. Crys., 43, 189-195.
[220] Ripmeester, J.A. Tse, J.S. Ratcliffe, C.I. & Powell, B.M. (1987). A new clathrate
hydrate structure. Nature, 325, 135-136.
[221] Ripmeester, J.A. & Ratcliffe, C.I. (1988). Low-temperature cross-polarization/magic
angle spinning carbon-13 NMR of number solid methane hydrates: structure, cage
occupancy, and hydration. J. Phys. Chem., 92, 337-339.
[222] Ripmeester, J.A. & Ratcliffe, C.I. (1990). Xenon-129 NMR studies of clathrate
hydrates: new guests for structure II and structure H. J. Phys. Chem., 94 (25), 8773-
8776.
[223] Ripmeester, J.A. & Ratcliffe, C.I. Klug, D.D. Tse, J.S. (1994). Molecular perspectives
on structure and dynamics in clathrate hydrates. Ann. N.Y. Acad. Sci., 715, 161-176.
[224] Ripmeester, J.A. & Ratcliffe, C.I. (1999). On the contributions of NMR spectroscopy
to clathrate science. J. Structural Chemistry, 40(5), 654-662.
[225] Ripmeester, J.A. (2000). Hydrate Research – From Correlations to a Knowledge-
based Discipline. Ann. N.Y. Acad. Sci., 912, 1-16.
[226] Robinson, D.B. & Ng, H.-J. (1986). Hydrate Formation and Inhibition in Gas or Gas
Condensate Streams. Journal of Canadian Petroleum Technology, 25, 26-30.
[227] Ross, D.K. (2006). Hydrogen storage: The major technological barrier to the
development of hydrogen fuel cell cars. Vacuum, 80 (10), 1084-1089.
[228] Salamatin, A.N. Hondoh, T. Uchida, T. & Lipenkov, V.Y. (1998). Post-nucleation
conversion of an air bubble to clathrate air-hydrate crystal in ice. J. Cryst. Growth,
193, 197-218.
[229] Schicks, J.M. & Ripmeester, J.A. (2002). The Coexistence of Two Different Methane
Hydrate Phases under Moderate Pressure and Temperature Conditions: Kinetic versus
Thermodynamic Products. Angew. Chem. Int. Ed., 43, 3310–3313.
[230] Schicks, J.M. Naumann, R. Erzinger, J. Hester, K.C. Koh, C.A. & Sloan, E.D. Jr.
(2006). Phase transition in mixed gas hydrates: Experimental observations versus
calculated data. J. Phys. Chem. B, 110, 11468-11474.
[231] Schneider, G.R. & Farrar, J. (1968). U.S. Dept. of Interior, Rsch. Dev. Rpt No 292,
p.37, January.
[232] Schneider, S. H. (1990). The global warming debate: science or politics? Environ. Sci.
Technol., 24, 432.
[233] Schuth, F. (2005). Technology - Hydrogen and hydrates. Nature, 434 (7034), 712-
713.
[234] Selim, M.S. & Sloan, E.D. (1989). Heat and mass transfer during dissociation of
hydrate in porous media. AIChE J., 35, 1049-1052.
54 Peter Englezos, John Ripmeester and Robin Susilo
[235] Selleck, F.T. Carmichael, L.T. & Sage, B.H. (1952). Phase Behavior in the Hydrogen
Sulfide-Water System. Ind. Eng. Chem., 44, 2219-2226.
[236] Seo, Y.-T. Kang, S.-P. & Lee, H. (2001). Experimental determination and
thermodynamic modeling of methane and nitrogen hydrates in the presence of THF,
propylene oxide, 1,4-dioxane and acetone. Fluid Phase Equilibria, 189, 99–110.
[237] Seo, Y.T. & Lee, H. (2003). 13C NMR Analysis and Gas Uptake Measurements of
Pure and Mixed Gas Hydrates: Development of Natural Gas Transport and Storage
Method using Gas Hydrate. Korean J. Chem. Eng., 20(6), 1085-1091.
[238] Seo, Y.T. Moudrakovski, I.L. Ripmeester, J.A. Lee, J.-W. & Lee, H. (2005a).
Efficient Recovery of CO2 from Flue Gas by Clathrate Hydrate Formation in Porous
Silica Gels. Environ. Sci. Technol., 39, 2315-2319.
[239] Seo, Y.-T. Kim, D.Y. Lee, H. Lee, J. Moudrakovski, I.L. & Ripmeester, J.A. (2005b).
Molecular behavior and guest distribution in the cages of CH4 + THF double hydrate.
Proc. Fifth Int. Con. Gas Hydrates, Vol. II, 568-572, Trondheim, Norway.
[240] Servio, P. & Englezos, P. (2003a). Morphology of methane and carbon dioxide
hydrates formed from water droplets. AIChE J., 49 (1), 269-276.
[241] Servio, P. & Englezos, P. (2003b). Morphology study of structure H hydrate
formation from water droplets. Crystal Growth & Design, 3 (1), 61-66.
[242] Servio, P. Macchi, A. & Luiz, D.S. (2004). Formation of gas hydrates by fluidized-
bed granulation. U.S. Pat. Appl. Publ..
[243] Shimada, W. Takeya, S. Kamata, Y. Uchida, T. Nagao, J. Ebinuma, T. & Narita, H.
(2005). Texture Change of Ice on Anomalously Preserved Methane Clathrate Hydrate.
J. Phys. Chem. B, 109, 5802-5807.
[244] Shpakov, V.P. Tse, J.S. Tulk, C.A. Kvamme, B. & Belosludov, V.R. (1998). Elastic
moduli calculation and instability in structure I methane clathrate hydrate. Chemical
Physics Letter, 282, 107–114.
[245] Simpson, S. (2000). Methane Fever. Scientific American, 2, 24.
[246] Sloan, E.D. Jr. & Fleyfel, F. (1992). Hydrate dissociation enthalpy and guest size.
Fluid Phase Equilibria, 76, 123-140.
[247] Sloan, E.D. Jr. (1998). Clathrate Hydrates of Natural Gasses, 2nd edition, New York,
NY: Marcel Dekker.
[248] Sloan, E.D. Jr. (2003a). Fundamental principles and applications of natural gas
hydrates. Nature, 426, 353-359.
[249] Sloan, E.D. Jr. (2003b). Clathrate hydrate measurements: microscopic, mesoscopic,
and macroscopic. J. Chem. Thermodynamics, 35, 41–53.
[250] Sloan, E.D. (2004a). Introductory overview: Hydrate knowledge development.
American Mineralogist, 89, 1155–1161.
[251] Sloan, E.D. (2004b). Thermodynamics of Natural Gas Hydrates, in Chemical
Thermodynamics for Industry, Fletcher, T.M., (editor), RSC, chapter 6.
[252] Sloan, E.D. (2005). A changing hydrate paradigm—from apprehension to avoidance
to risk management. Fluid Phase Equilibria, 228, 67-74.
[253] Stackelberg, M.V. & Müller, H.R. (1951). Zur struktur der gashydrate. Naturwiss, 38,
456.
[254] Staykova, D.K. Kuhs, W.F. Salamatin, A.N. & Hansen, T. (2003). Formation of
Porous Gas Hydrates from Ice Powders: Diffraction Experiments and Multistage
Model. J. Phys. Chem. B, 107 (37), 10299-10311.
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 55
[255] Stern, L.A. Circone, S. Kirby, S.H. & Durham, D.B. (2003). Temperature, pressure,
and compositional effects on anomalous or "self" preservation of gas hydrates. Can. J.
Phys., 81 (1-2), 271-283.
[256] Stern, L.A. Kirby, S.H. & Durham, W.B. (1998). Polycrystalline Methane Hydrate:
Synthesis from Superheated Ice, and Low-Temperature Mechanical Properties.
Energy and Fuels, 12 (2), 201-211.
[257] Stern, L.A. Circone, S. Kirby, S. H. & Durham, W.B. (2001a). Anomalous
Preservation of Pure Methane Hydrate at 1 atm. J. Phys. Chem. B, 105, 1756-1762.
[258] Stern, L.A. Circone, S. Kirby, S.H. & Durham, W.B. (2001b). Preservation of
methane hydrate at 1 Atm. Energy and Fuels, 15 (2), 499-501.
[259] Stern, L.A. Kirby, S.H. Circone, S. & Durham, W.B. (2004). Scanning Electron
Microscopy investigations of laboratory-grown gas clathrate hydrates. American
Mineralogist, 89, 1162–1175.
[260] Stevenson, D.J. (1999). Life-sustaining planets in interstellar space. Nature,
400(6939), 32.
[261] Strobel, T.A. Taylor, C.J. Hester, K.C. Dec, S.F. Koh, C.A. Miller, K.T. &Sloan, E.D.
Jr. (2006). Molecular hydrogen storage in binary THF-H-2 clathrate hydrates. J. Phys.
Chem. B, 110(34), 17121-17125.
[262] Subramanian, S. Kini, R.A. Dec, S.F. & Sloan, E.D. Jr. (2000a). Evidence of structure
II hydrate formation from methane-ethane mixtures. Chem. Eng. Sci., 55, 1981-1999.
[263] Subramanian, S. Ballard, A.L. Kini, R.A. Dec, S.F. & Sloan, E.D. Jr. (2000b).
Structural transitions in methane-ethane gas hydrates, Part I: upper transition point
and applications. Chem. Eng. Sci., 55, 5763-5771.
[264] Suess, E. Bohrmann, G. Greinert, J. & Lausch, E. (1999). Flammable Ice. Scientific
American, 11, 76.
[265] Sultan, N. Cochonat, P. Foucher, J.-P. & Mienert, J. (2004a). Effect of gas hydrates
melting on seafloor slope instability. Marine Geology, 213 (1-4), 379-401.
[266] Sultan, N. Cochonat, P. Canals, M. Cattaneo, A. Dennielou, B. Haflidason, H. Laberg,
J.S. Long, D. Mienert, J. Trincardi, F. Urgeles, R. Vorren, T.O. & Wilson, C. (2004b).
Triggering mechanisms of slope instability processes and sediment failures on
continental margins: a geotechnical approach. Marine Geology, 213 (1-4), 291-321.
[267] Sum, A.K. Burrus, R.C. & Sloan, E.D. Jr. (1997). Measurement of clathrate hydrates
via Raman spectroscopy. J. Phys. Chem. B, 101 (38), 7371-7377.
[268] Sun, Z.-G. Fan, S.-S. Guo, K.-H. Shi, L. & Wang, R.-Z. (2002). Equilibrium hydrate
formation conditions for methylcyclohexane with methane and a ternary gas mixture.
Fluid Phase Equilibria, 198, 293-298.
[269] Sun, Z. Wang, R. Ma, R. Guo, K. & Fan, S. (2003). Effect of surfactants and liquid
hydrocarbons on gas hydrate formation rate and storage capacity. Int. J. Energy Res.,
27, 747-756.
[270] Sun, C.-Y. & Chen, G.-J. (2006). Methane hydrate dissociation above O degrees C
and below O degrees C. Fluid Phase Equilibria, 242 (2), 123-128.
[271] Susilo, R. Lee, J.D. & Englezos, P. (2005). Liquid–liquid equilibrium data of water
with neohexane, methylcyclohexane, tert-butyl methyl ether, n-heptane and vapor–
liquid–liquid equilibrium with methane. Fluid Phase Equilibria, 231, 20-26.
56 Peter Englezos, John Ripmeester and Robin Susilo
[272] Susilo, R. Moudrakovski, I.L. Ripmeester, J.A. & Englezos, P. (2006). Hydrate
kinetics study in the presence of non-aqueous liquid by NMR spectroscopy and
imaging. J. Phys. Chem. B, 110 (51), 25803-25909.
[273] Susilo, R. Ripmeester, J.A. & Englezos, P. (2007a). Characterization of gas hydrates
with PXRD, DSC, NMR and Raman Spectroscopy, accepted for publication in Chem.
Eng. Sci..
[274] Susilo, R. Ripmeester, J.A. & Englezos, P. (2007b). Methane uptake and conversion
rate into structure H hydrate, submitted for publication to AIChE J.
[275] Svandal, A. Kvamme, B. Granasy, L. Pusztai, T. Buanes, T. & Hove, J. (2006). The
phase-field theory applied to CO2 and CH4 hydrate. J. Crystal Growth 287, 486-490.
[276] Tajima, H. Yamasaki, A. & Kiyono, F. (2004). Continuous formation of CO2 hydrate
via a kenics-type static mixer. Energy and Fuels, 18 (5), 1451-1456.
[277] Tajima, H. Yamasaki, A. & Kiyono, F. (2005a). Effects of mixing functions of static
mixers on the formation of CO2 hydrate from the two-phase flow of liquid CO2 and
water. Energy and Fuels, 19 (6), 2364-2370.
[278] Tajima, H. Yamasaki, A. & Kiyono, F. (2005b). Process design of a new injection
method of liquid CO2 at the intermediate depths in the ocean using a static mixer.
Fuel Processing Technology, 86 (14-15), 1667-1678.
[279] Tajima, H. Yamasaki, A. Kiyono, F. & Teng, H. (2006). Size distribution of CO2
drops in a static mixer for ocean disposal. AIChE J., 52 (8), 2991-2996.
[280] Takahashi, M. Kawamura, T. Yamamoto, Y. Ohnari, H. Himuro, S. & Shakutsui, H.
(2003). Effect of Shrinking Microbubble on Gas Hydrate Formation. J. Phys. Chem.
B, 107 (10), 2171-2173.
[281] Takeya, S. Shimada, W. Kamata, Y. Ebinuma, T. Uchida, T. Nagao, J. & Narita, H.
(2001). In Situ X-ray Diffraction Measurements of the Self-Preservation Effect of
CH4 Hydrate. J. Phys. Chem. A, 105, 9756-9759.
[282] Takeya, S. Hondoh, T. & Uchida, T. (2002). Insitu observations of CO2 hydrate by
X-Ray diffraction. Ann. NY Acad. Sci., 912, 973-982.
[283] Takeya, S. Kamata, Y. Uchida, T. Nagao, J. Ebinuma, T. Narita, H. Hori, A. &
Hondoh, T. (2003). Coexistence of structure I and II hydrates formed from a mixture
of methane and ethane gases. Can. J. Phys., 81, 479–484.
[284] Takeya, S. Uchida, T. Nagao, J. Ohmura, R. Shimada, W. Kamata, Y. Ebinuma, T. &
Narita, H. (2005). Particle size effect ofCH4 hydrate for self-preservation. Chem. Eng.
Sci., 60, 1383-1387.
[285] Takeya, S. Kida, M. Minami, H. Sakagami, H. Hachikubo, A. Takahashi, N. Shoji, H.
Soloviev, V. Wallmann, K. Biebow, N. Obzhirov, A. Salomatin, A. & Poort, J.
(2006). Structure and thermal expansion of natural gas clathrate hydrates. Chem. Eng.
Sci., 61, 2670-2674.
[286] Tanasawa, I. & Takao, S. (2002). Clathrate hydrate slurry of tetra-n-butylammonium
bromide as a cold storage material. Proc. 4th Int. Conf. Gas Hydrates, 963-967,
Yokohama, Japan.
[287] Taylor, F. W. (1991). The greenhouse effect and climate change. Reports on Progress
in Physics, 54, 881.
[288] Taylor, C.E. (2005). Photocatalytic conversion of methane contained in methane
hydrates. Topics in Catalysis, 32 (3-4), 179-184.
Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 57
[289] Ternes, M.P. (1984). Characterization of refrigerant - 12 gas hydrate formation for
heat pump cool storage applications. ASME energy sources technology conference,
84-AES-12, American Society of Mechanical Engineers, New York, USA.
[290] Thomas, S. & Dawe, R.A. (2003). Review of ways to transport natural gas energy
from countries which do not need the gas for domestic use. Energy, 28 (14), 1461-
1477.
[291] Thompson, H. Soper. A.K. Buchanan, P. Aldiwan, N. Creek, J.L. & Koh, C.A.
(2006). Methane hydrate formation and decomposition: Structural studies via neutron
diffraction and empirical potential structure refinement. J. Chem. Phys., 124 (16), Art.
No. 164508.
[292] Tobie, G. Lunine, J.I. & Sotin, C. (2006). Episodic outgassing as the origin of
atmospheric methane on Titan. Nature, 440 (7080), 61-64.
[293] Tohidi, B. Danesh, A. Todd, A.C. Burgass, R.W. Østergaard, K.K. (1996).
Equilibrium data and thermodynamic modelling of cyclohexane gas hydrates. Chem.
Eng. Sci. 51(1), 159-163.
[294] Tohidi, B. Danesh, A. Burgass, R.W. Todd, A.C. Østergaard, K.K. (1997).
Equilibrium data and thermodynamic modelling of cyclopentane and neopentane
hydrates. Fluid Phase Equilibria, 138, 241-250.
[295] Tombari, E. Presto, S. Salvetti, G. & Johari, G.P. (2006). Heat capacity of
tetrahydrofuran clathrate hydrate and of its components, and the clathrate formation
from supercooled melt. J. Chem. Phys., 124 (15), 154507.
[296] Tomlison, J.J. (1982). Heat pump cool storage in a clathrate of Freon. Proc. 17th
Intersoc Energy Conversion Eng. Conf., Vol.4, 2060-2064, New York, USA.
[297] Trebble, M. A. & Bishnoi P. R. (1988). Extension of the Trebble-Bishnoi Equation of
State to Fluid Mixtures. Fluid Phase Equilib., 40, 1-21.
[298] Tse, J.S. Handa, Y.P. Ratcliffe, C.I. & Powell, B.M. (1986). Structure of oxygen
clathrate hydrate by neutron powder diffraction. J. Incl. Phen. Microcyclic Chem., 4,
235-240.
[299] Tsuji, H. Ohmura, R. & Mori, Y.H. (2004). Forming Structure-H Hydrates Using
Water Spraying in Methane Gas: Effects of Chemical Species of Large-Molecule
Guest Substances. Energy and Fuels, 18(2), 418-424.
[300] Tsuji, H. Kobayashi, T. Ohmura, R. & Mori, Y.H. (2005a). Hydrate Formation by
Water Spraying in a Methane + Ethane + Propane Gas Mixture: An Attempt at
Promoting Hydrate Formation Utilizing LMGS for sH Hydrates. Energy and Fuel, 19,
869-876.
[301] Tsuji, H. Kobayashi, T. Okano, Y. Ohmura, R. Yasuoka, K. & Mori, Y.H. (2005b).
Thermodynamic Simulations of Isobaric Hydrate-Forming Operations: Formulation
of Computational Scheme and Its Application to Hydrate Formation from a Methane
+ Ethane + Propane Mixture. Energy and Fuel, 19, 1587-1597.
[302] Tulk, C.A. Ripmeester, J.A. & Klug, D.D. (2000). The application of Raman
spectroscopy to the study of gas hydrates. Ann. N.Y. Acad. Sci., 912, 859-872.
[303] Uchida, T. Hondoh, T. Mae, S. Lipenkov, V.Y. & Duval, P. (1994). Air-hydrate
crystals in deep ice-core samples from Vostok Station Antartica. J. Glaciology, 40
(134), 79-88.
58 Peter Englezos, John Ripmeester and Robin Susilo
[320] Yeon, S.-H. Seol, J. & Lee, H. (2006). Structure Transition and Swapping Pattern of
Clathrate Hydrates Driven by External Guest Molecules. J. Am. Chem. Soc., 128(38),
12388-12389.
[321] Yoon, J.-H. Yamamoto, Y. Komai, T. Haneda, H. (2003). Rigorous Approach to the
Prediction of the Heat of Dissociation of Gas Hydrates. Ind. Eng. Chem. Res., 42,
1111-1114.
[322] Yoon, J.-H. Kawamura, T. Yamamoto, Y. & Komai, T. (2004). Transformation of
Methane Hydrate to Carbon Dioxide Hydrate: In Situ Raman Spectroscopic
Observations. J. Phys. Chem. A, 108, 5057-5059.
[323] Yoon, J.-H. Kawamura, T. Ohtake, M. Takeya, S. Komai, T. Yamamoto, Y. Emi, H.
Kohara, M. Tanaka, S. Takano, O. & Uchida, K. (2006). Highly selective encaging of
carbon dioxide molecules in the mixed carbon dioxide and nitrogen hydrate at low
temperatures. J. Phys. Chem. B, 110 (35), 17595-17599.
[324] Zachos, J. C. Lohmann, K. C. Walker, J. C. G. & Wise, S. W. (1993). Abrupt climate
change and transient climates during the paleocene: a marine perspective. Journal of
Geology, 101, 191.
[325] Zeng, H. Wilson, L.D. Walker, V.K. & Ripmeester, J.A. (2003). The inhibition of
tetrahydrofuran clathrate-hydrate formation with antifreeze protein. Can. J. Phys., 81
(1-2), 17-24.
[326] Zeng, H. Wilson, L.D. Walker, V.K. & Ripmeester, J.A. (2006a). Effect of antifreeze
proteins on the nucleation, growth, and the memory effect during tetrahydrofuran
clathrate hydrate formation. J. Am. Chem. Soc., 128 (9), 2844-2850.
[327] Zeng, H. Moudrakovski, I.L. Ripmeester, J.A. & Walker, V.K. (2006b). Effect of
antifreeze protein on nucleation, growth and memory of gas hydrates. AIChE J., 52
(9), 3304-3309.
[328] Zhong, Y. & Rogers, R.E. (2000). Surfactant effects on gas hydrate formation.
Chem. Eng. Sci., 55, 4175-4187.
In: Electroanalytical Chemistry: New Research ISBN: 978-1-60456-347-4
Editor: G. M. Smithe © 2008 Nova Science Publishers, Inc.
Chapter 2
Abstract
Atmospheric corrosion is the most extended type of corrosion in the World. Over the years,
several papers have been published in this subject; however, most of the research has been
made in non-tropical countries and under outdoor conditions. Results of outdoor and indoor
corrosion rate and corrosion aggressivity in tropical corrosion test stations of Cuba and
Mexico are reported.
Time of wetness (TOW), considered as the time during which the corrosion process occurs, is
an important parameter to study the atmospheric corrosion of metals. According to ISO-9223
standard, TOW is approximately the time when relative humidity exceeds 80% and
temperature is higher than 0oC. No upper limit for temperature is established. In tropical
climates, when temperature reaches values over 25oC, evaporation of water plays an important
role and the possibility to establish an upper limit respecting temperature should be analyzed.
The concept of TOW assumes the presence on the metallic surface of a water layer; however,
there are recent reports about the formation of water microdrops during the initial periods of
atmospheric corrosion, showing that the idea of the presence of thin uniform water layers is
not completely in agreement with the real situation in some cases (particularly indoor
exposures).
62 F. Corvo, T. Pérez, Y. Martin et al.
Most of the research carried out to study the initial stages of atmospheric corrosion have been
made on a clean surface without corrosion products; however, the metal is very often covered
by thin or thick corrosion products after a given exposure time and these products usually act
as retarders of the corrosion process.
In the Cuban Isle, the influence of chloride ions is very significant in determining the
corrosion rate. In the coastal territory of the Mexican Gulf, particularly at Campeche, the
deposition of Chloride ions is lower. No previous reports have been made about the
interaction between chloride deposition rate and rain. The influence of rain seems to be
important in determining the acceleration rate of chloride ions on metals due to its washing
effect. To consider the influence of the interaction chloride deposition rate–rain regime could
be useful to improve the prognosis of corrosion aggressivity. The predominant wind direction
corresponding to geographic sites result in an important parameter for chloride deposition and
their influence on surface wetness.
The calculation of Time of Wetness established in ISO 9223 should be revised based on new
results obtained in outdoor and indoor conditions in tropical humid marine climate. Some
proposals are made to improve the estimation of TOW, taking into account changes in its
nature depending on outdoor or indoor exposure, linear relationship between time and TOW,
the effect of rain, and the role of contaminants and air temperature.
Introduction
Atmospheric corrosion is the most extended type of corrosion in the World. Over the
years, several papers have been published in this subject; however, most of the research has
been made in non-tropical countries and under outdoor conditions. The tropical climate is
typical of equatorial and tropical regions and is characterized by permanently high
temperatures and relative humidity with considerable precipitation, at least during part of the
year. A high corrosion rate of metals is usually reported for this climate.
Results of outdoor and indoor corrosion rate and corrosion aggressivity in tropical
corrosion test stations of Cuba and Mexico are reported. The results mainly concern to natural
atmospheric corrosion tests obtained in the western side of the Isle of Cuba and in the
Campeche State located at the Yucatán Peninsula in Mexico. The two regions are located in
the tropical climate and receive the influences of the waters of the Atlantic Ocean and the
Mexican Gulf. Data processed in this paper correspond to atmospheric corrosion tests carried
out during a long period of time, about the last 20 years in Cuba and the last 10 years at
Campeche up to the present.
transportation. Depending on the contaminants sources existing in different regions, the kind
of contaminants may change, but in general, the main types of contaminants are Chlorides
and Sulfur compounds
Air temperature in the Yucatán Peninsula reaches higher values than in Cuba, the top
average air temperature in Cuba is 26oC in the eastern shoreline [7]. In the Yucatán Peninsula
part of the territory can be classified as Tropical very warm with an average temperature over
26oC [8]. Absolute maximum temperature in Cuba is of 38,6oC, meanwhile a temperature
over 40oC is frequently reported in sites of the Yucatán Peninsula, particularly in Campeche.
About ¼ of the Cuban territory is composed by mountains and the rest is mainly terrain
flatness. The Yucatán Peninsula has no mountains. The lower temperatures and higher
precipitations rates in Cuba are localized in the mountain regions. Orography is a
transformation factor of local wind regime.
Average relative humidity in the west side of Cuba is frequently higher than in
Campeche, the average is about 80% for Cuba and about 76% for Campeche. Daily maximum
relative humidity values are over 90% for the west part of Cuba and over 88% for Campeche.
In spite of this, the classification of TOW according to ISO-9223[9] is the same for the two
territories: τ4, corresponding to “Outdoor atmospheres at all climates, excepting dry and cold
climates”.
The influence of air temperature may change depending on the conditions. The effect of
increasing temperature appears to be an increase in corrosion rate under conditions of
permanent surface wetting, such as those obtained during precipitation. Under conditions of
varying surface wetting, however, the corrosion rate increases with temperature up to a
certain maximum value, and thereafter decreases [10]. ISO 9223 standard established that for
purposes of aggressivity determinations the use of temperature-humidity complex data is the
recommended methodology, because corrosion rate depends on this complex and not only on
temperature or humidity independently. Taking into account this report, a decrease in
corrosion rate should be expected in tropical climate for an increase in average air
temperature.
[11] defines: The wetting of surfaces is caused by many factors, for example, dew, rainfall,
melting snow and a high humidity level. The length of time when the relative humidity is
greater than 80% at a temperature greater than 00C is used to estimate the calculated time of
wetness (τ) of corroding surfaces. Information on calculated time of wetness is helpful for
informative atmosphere corrosivity estimation. In this standard, TOW is “estimated” based on
the characteristics of the temperature humidity complex, independently of the pollutant level
and the nature of the metal or alloy. It is not the actual time of wetness, but it is an estimation
based only on climatic factors (temperature and relative humidity) and independent of the
nature and characteristics of the metallic surface. Some results obtained in the tropical regions
of the Gulf of Mexico suggest that an upper limit in temperature for the definition of TOW
could be established [3-4,12-14]. TOW as defined in ISO 9223 does not cover all the aspects
of climate. According to the ISO definition, in the case of tropical climate where air
temperature never reach 0oC, TOW is estimated as the time when relative humidity is over
80%. The diminition of the electrolyte layer at relative humidity over 80% when temperature
is over 25oC has been reported for Santiago de las Vegas rural station in Cuba [12] using
TOW sensors based in a circuit print covered with a gold layer. By other way, metal
temperature and TOW were registered in a rural station at 30 km of the seashore located in
Merida, Yucatan Peninsula, using copper/gold sensors [14] according to ASTM Practice G84-
89 (1999). It was reported the influence of the nature of the metal and their corrosion
products, the orientation of the sample (skyward or groundward), the direction of the winds
and rain precipitation on the measured TOW values. None of this factor is taken into
consideration by ISO 9223. A quite different temperature was determined for the metal
surface respecting air temperature. It is perhaps an explanation to the very possible increase in
evaporation of the surface electrolyte when air temperature is over 25oC.
Water adsorption on silver surfaces exposed into a ventilated shed in an urban-rural site
of Cuba was studied [13] using quartz resonators covered with a silver layer. It was
determined that in these indoor conditions water adsorption significantly diminishes when air
temperature increases over 25oC at relative humidity ranges of 80-90% and 90-100%. All
these results confirm the idea that an upper limit of temperature should be established for the
estimation of time of wetness.
The presence of water does not only create conditions for the existence of an electrolyte,
but it acts as a solvent for the dissolution of contaminants [10]. Oxygen plays an important
role as oxidant element in the atmospheric corrosion process. The thickness of the water layer
determines the oxygen diffusion toward the metallic surface and also the diffusion of the
reaction products to the outside interface limited by the atmosphere. Another aspect of ISO
definition is that “a metallic surface is covered by adsorptive and/or liquid films of
electrolyte”. According to new results, the presence of adsorptive or liquid films of electrolyte
perhaps could be not in the entire metallic surface, but in places where there is formed a
central anodic drop due to the existence of hygroscopic particles or substances surrounded by
microdrops where the cathodic process takes place. This phenomenon is particularly possible
in indoor conditions [15-18].
Using ISOCORRAG, MICAT and Russian data, Tidblad et al [19] showed that the
inclusion of temperature among the environmental parameters improves considerably the
usefulness of the dose-response functions and should be adapted in the revision of ISO 9223
standard. It is reported an increase in corrosion rate with average air temperature in the range
of -15 to 30oC.
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 65
In the eighties of the last century the adsorption of water layers on surfaces was studied
[20]. It was demonstrated that the change in thickness of these layers depends on the physico-
chemical properties of water in these thin water layers. It is reported that on iron surfaces, the
number of adsorbed water layers is about 15 at RH 55% and 90 at 100%. Similar values are
obtained for Copper and Zinc; however significant differences are reported for Platinum,
gold, aluminum and silver. These monolayers have been calculated only in presence of water
(without oxygen) where the corrosion process is very slow and, consequently, in conditions
far from the reality.
Almost all tests carried out to study the starting process of atmospheric corrosion have
been performed in a surface without corrosion products; however, in real conditions, the
metal is covered with corrosion products after a given time and these products begin to play
its role as retarders of the corrosion process in almost all cases. Corrosion products acts as a
barrier for oxygen and contaminants diffusion, the free area for the occurrence of the
corrosion is lower; however, the formation of the surface electrolyte is enhanced. Only in
very polluted areas the corrosion products accelerate the corrosion process. Water adsorption
isoterms were determined to corrosion products formed in Cuban natural atmospheres[21].
Sorption properties of corrosion products (taking into account their salt content-usually
hygroscopics) determine the possibilities of surface adsorption and the possibility of
development of corrosion process
For a single type of climate could be more practical and easy the use of time instead of
TOW, because it is expected the same category of TOW according to ISO 9223. In other
words, it makes no sense to work using TOW-ISO in tropical climate, because in general it is
in the same category at all places, so its influence is the same. Using data for different
climates or wide regions a significant change in TOW-ISO could take place, including
changes in wet/dry cycles. In these conditions it should be recommendable to use TOW
according to ISO definition.
Outdoor TOW-ISO
Corrosion rate is a function of time of wetness, considered as the time during which
corrosion occurs, but in general it should not be a linear function because corrosion rate
changes with time. There are different factors influencing, for example, the protective
properties of the corrosion products, the increase or decrease of the acceleration caused by
contaminants, increase or decrease of the thickness and conductivity of the electrolyte layer,
i.e.
If the definition of TOW established by ISO is used (TOW-ISO), a linear relationship
between time and TOW is obtained, in spite of the different possible changes in corrosion rate
caused by changes in the nature of TOW. It has to be remarked that it is not the same effect
on corrosion rate caused by a heavy rain than dew, fog or water adsorption, so for the same
interval of relative humidity (80-100%), notable changes in nature of TOW-ISO and
consequently in corrosion rate could take place.
If a linear regression is fitted between time (sum of time in months) and TOW-ISO (sum
of TOW corresponding to every month up to one year), a perfect linear relationship is
obtained (see tables I, II, III, IV and V). The results show that a perfect linear relation exists
66 F. Corvo, T. Pérez, Y. Martin et al.
between time and TOW-ISO. It means that both variables are equivalent, so for purposes of
long term prognosis, work carried out time give the same result than using TOW-ISO. It does
not mean that the actual time of wetness should be the same as estimated according to ISO,
because as it has been pointed out above, the ISO definition does not take into account the
nature of the different components of TOW regarding climate (rain, dew, fog, water
adsorption) and the nature of the metal and the corrosion products.
Table I. Simple regression of TOW-ISO data and time for Campeche PCGM coastal test
station in a period of 2 years. Equation t = a + b TOW-ISO, r = correlation coefficient,
r2 = percentage of variation explained by the independent variable, P = Statistical
Probability
Year a b r r2 P
2004 -230.42 234.39 0.979 95.90 <0.0000
2005 -300.15 258.77 0.974 94.88 <0.0000
Campeche PCGM
Table II. Simple regression of TOW-ISO data and time for Viriato coastal test station in
a period of 10 years. Equation t = a + b TOW-ISO. , r = correlation coefficient, r2 =
percentage of variation explained by the independent variable, P = Statistical
Probability
Year a b r r2 P
1987 -487.20 406.48 0.987 97.49 <0.0000
1988 -14.83 423.15 0.997 99.38 <0.0000
1989 -160.20 365.59 0.998 99.58 <0.0000
1990 89.30 299.48 0.996 99.11 <0.0000
1991 -167.38 395.20 0.998 99.52 <0.0000
1992 -237.26 375.55 0.996 99.16 <0.0000
1993 -91.18 348.37 0.996 99.29 <0.0000
1994 -70.60 380.73 0.999 99.71 <0.0000
1995 -428.05 413.16 0.994 98.73 <0.0000
1996 83.97 252.75 0.997 99.489 <0.0000
1997 -101.30 390.60 0.999 99.73 <0.0000
Table III. Simple regression of TOW-ISO data and time for Holguin coastal station
(located in the eastern side of Cuba) in a period of 3 years.
Year a b r r2 P
1987 126.42 469.84 0.999 99.79 <0.0000
1988 27.13 435.98 0.999 99.88 <0.0000
1989 54.44 403.36 0.999 99.96 <0.0000
Equation t = a + b TOW-ISO. , r = correlation coefficient, r2 = percentage of variation explained by the
independent variable, P = Statistical Probability
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 67
Table IV. Simple regression of TOW-ISO data and time for Cojimar coastal station in a
period of 10 years. Equation t = a + b TOW-ISO. , r = correlation coefficient, r2 =
percentage of variation explained by the independent variable, P = Statistical
Probability
Year a b r r2 P
1991 -127.68 389.13 0.998 99.52 <0.0000
1992 -172.08 367.73 0.996 99.14 <0.0000
1993 -54.0 342.23 0.998 99.53 <0.0000
1994 -51.32 381.34 0.999 99.78 <0.0000
1995 -289.58 389.58 0.993 98.60 <0.0000
1996 -164.67 376.75 0.997 99.41 <0.0000
1997 -51.44 369.95 0.999 99.83 <0.0000
1998 -75.36 344.79 0.997 99.40 <0.0000
1999 -286.80 352.04 0.989 97.89 <0.0000
2000 -159.21 371.14 0.995 99.02 <0.0000
Table V. Simple regression of TOW-ISO data and time for Quivican rural station in a
period of 8 years. Equation t = a + b TOW-ISO. , r = correlation coefficient, r2 =
percentage of variation explained by the independent variable, P = Statistical
Probability
Year a b r r2 P
1990 -89.37 405.78 0.997 99.34 <0.0000
1991 -121.50 421.15 0.997 99.34 <0.0000
1992 -66.07 440.34 0.998 99.59 <0.0000
1993 -33.53 443.97 0.999 99.85 <0.0000
1994 -75.86 448.22 0.999 99.73 <0.0000
1995 -105.04 458.38 0.999 99.72 <0.0000
1996 -62.50 444.04 0.999 99.53 <0.0000
1997 -71.86 457.37 0.999 99.78 <0.0000
The results show that a perfect relation exists between time and TOW-ISO, so both
variables are equivalent and can be used for purposes of long term prognosis.
There are some small differences between the metallic boxes used in Cuba and
Campeche. On figures 1 and 2 it can be observed that Cuban metallic boxes are relatively
more closed than the one used at Campeche PCGM corrosion station.
Figure 1. Metallic box exposed at Cojimar coastal station in the north shore of the City of Havana.
Similar metallic boxes are exposed at CNIC and Quivican Cuban stations. Samples and contaminant
detector are placed inside the metallic boxes.
Figure 2. Metallic box exposed at Campeche PCGM coastal station at 300 m of the shoreline in the City
of Campeche (front view). Samples and contaminant detectors are placed inside the metallic boxes. As
it can be observed, this metallic box has a wider window than those used in Cuban stations.
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 69
K=aτb
Where:
a = constant
b = coefficient that indicates the protective properties of the corrosion products layer
τ = Actual time of wetness (effective time during which corrosion process occurs)
It is very wide used in studying the atmospheric corrosion process and its development
with time. As smaller will be coefficient b, it is supposed that the protective properties of the
corrosion products layer are higher.
On Table VI it is shown a comparison between data fitness using TOW calculated
according to ISO 9223 and time. It can be observed that there are not significant differences
in using TOW or time because r2 is almost the same in both cases. It could be explained
supposing a direct relationship between time and TOW, that is, TOW increases with time. If,
in general, there should be differences in TOW for the different climatic seasons, these are
not significant for the total time.
Table VI. Statistical fitness of data of corrosion of Copper, Steel, Nickel and Tin inside a
metallic box (heat trap conditions) respecting time and TOW according to ISO
definition.
K=aτb K= a t b
Station Metal 2
a b r a b r2
Cojimar Copper 0.0009 0.847 48.93 0.138 0.805 49.76
(coastal) Steel 0.0038 1.185 96.27 4.284 1.455 97.11
Tin 0.0014 0.817 88.09 0.177 0.795 98.41
Nickel n.w.l. n.w.l. n.w.l. n.w.l. n.w.l. n.w.l.
CNIC Copper 1.1270 1.239 90.85 0.205 1.179 91.76
(urban) Steel 1.7930 1.596 94.39 2.823 1.532 95.79
Tin 0.0044 0.766 85.15 0.467 0.716 82.98
Nickel 9.7180 1.216 97.54 0.015 1.163 95.98
Quivicán Copper 0.0017 0.986 98.31 0.453 0.976 97.81
(rural) Steel 0.0206 1.087 97.83 9.412 1.083 98.68
Tin 0.0135 0.705 70.76 0.728 0.697 70.30
Nickel 7.2561 0.953 94.82 0.158 0.945 94.67
Campeche Copper 0.00003 1.528 93.68 0.158 1.628 90.68
PCGM Steel 0.0080 1.137 93.33 4.141 1.209 90.84
(coastal) Tin 0.0022 0.914 88.23 0.354 0.975 85.62
Nickel 0.0000002 1.851 88.25 0.006 1.974 85.55
n.w.l.= No weight loss detected
70 F. Corvo, T. Pérez, Y. Martin et al.
Indoor Humidity
Under indoor conditions, in the same way than outdoors, it is necessary the presence of
surface humidity for corrosion to occur due to the electrochemical nature of the atmospheric
corrosion process; however, in indoor conditions there are no precipitations and the presence
of surface water depends mainly on water content in the air and changes in temperature on the
surface, as well as the presence of hygroscopic substances on the metallic surface.
Indoor temperature and relative humidity appreciably depends on the ventilation level, on
the use or not of air conditioner or heating systems and of the thermal isolation. Leygraf
reports[23] that indoor relative humidity ranges from 15 to 85% with an average value of
50%. In Cuba [24] average values significantly different to 50% have been found, because
inside a closed storehouse (no windows) during one year an average value of 89,5% is
reported. In a storehouse having windows the average value is of 79,5%. In a storehouse
having a dehumidifier system the annual average relative humidity diminishes to 74,5%, still
significantly high respecting Leygraf report. In two storehouses without climatic control it is
reached a 100% of relative humidity, while in the storehouse submitted to dehumidification
the relative humidity reaches 85%. Changes in indoor relative humidity are lower than those
occurring outdoors, without the strong change caused by the difference between night and
day. In this way, the most probable way the metallic surface can be humid is through water
adsorption, although it can be observed that a significant relative humidity is reported indoors
respecting Europe, so a higher corrosion should be determined. The higher temperature of the
Cuban climate causes also the existence of higher water content and increases the possibilities
of water condensation.
Under indoor conditions, wind rate is lower than outdoors [25]. In this way, outdoor
conditions promotes a fast surface drying due to the influence of air and sun radiation, and
also a more frequent surface wetness thicker water layers due to rainfall, factors almost
nonexistent in indoor conditions. Another factor nonexistent in indoor conditions is the
washing effect on the metallic surface by precipitations.
It is important to remark that indoors, TOW-ISO is higher in storehouses without
ventilation than outdoors in tropical humid climates. ISO 9223 classifies TOW for no
ventilated storehouses as τ5 (more than 5500 h/a), whereas for outdoor conditions it is τ4
(2500-5500 h/a). Taking into account that indoors corrosion rate is significantly lower
respecting outdoors, it can be perfectly understood that the development of corrosion under
adsorbed water is very low respecting outdoors. In these conditions, the magnitude of
corrosion will be determined by the possibilities of acceleration that could cause
contaminants, being the presence of water on the surface a requirement for the occurrence of
corrosion, but not for determining the rate of the corrosion reaction.
When the wind velocity is lower, drying process could be larger and could be a
significant part of the dry/wetness cycle. It also depends on the temperature existing on the
metallic surface. On ISO-9223 there are not considered changes in the dry/wet cycles.
The deposition of contaminants in the surface usually causes an acceleration of the
corrosion process; however, there should not be excluded the possibility that a given
contaminant could diminish corrosion rate, as it could be the case of ammonia and its
influence on steel (due to its alkaline properties, it could induce the passivation of steel).
An important difference between outdoor and indoor conditions is that in the last one
there is no washing of contaminants by precipitations. Contaminants deposition is lower, but
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 71
it increases with time. In these conditions, the presence of adsorbed water is necessary for
corrosion to occur, but the rate of the process depends on the acceleration cause by
contaminants. The calculation of TOW-ISO should not be as important as outdoors, because
adsorbed water could be present all time. It has been reported a significant decrease in
adsorbed water as temperature increases in Cuban climate [13].
In indoor conditions, the role of airborne salinity significantly diminishes, because the
deposition rate is significantly lower than outdoors [24].
Stratmann reports [26-27] polarization curves on metallic surfaces covered by electrolyte
layers as thin as 2 micrometers without using Luggin capillary, but using a Kelvin probe as
reference (it is not necessary a contact to the surface). In this way there is avoided the effect
in the polarization curve caused by the presence of a capillary having aqueous solution inside.
It was possible to study the drying process, observing that at the beginning oxygen reduction
increases due to a fast transport of this element through the thin electrolyte layer. Corrosion
rate was determined by the oxygen consumption in a closed volume. It was shown that
corrosion rate of pure steel could be divided into three parts: the increase in the oxygen
diffusion current at the beginning, passivity of the surface in a second period and the
diminishing in the oxygen reduction at the end of the drying process.
Recent reports about the microdroplets formation in the starting periods of atmospheric
corrosion [15-18] show that the idea of a thin uniform water layers is not completely in
accordance with the reality. It has been observed that when a water drop is on the metallic
surface, formed in the place where a salt deposit existed before, microdroplets are formed
around this central drop. The cathodic process takes place in these surrounding microdroplets,
meanwhile the anodic process takes place in the central drop. This idea is not consistent with
the proposal of an uniform water layer on the surface and it is very probable that this situation
could be obtained under indoor conditions. It has been determined that microdrops (about 1
micron diameter) clusters are formed around a central drop. An important influence of air
relative humidity is reported on microdrops formation. There is a critical value of relative
humidity for the formation of microdroplets. Under this value no microdroplets are formed.
This value could be considered as the critical relative humidity. This situation is very similar
to the process of indoor atmospheric corrosion: presence of humid air, deposition of
hygroscopic contaminants in the surface, formation of microdrops. Water is necessary for
corrosion reaction to occur, but the reaction rate depends on the deposition rate and nature of
contaminants.
In ISO 11844-3[28] it is established that the combination of different parameters is what
determines the corrosivity of the atmosphere. Under indoor conditions, corrosion process
depends on a more complex number of parameters than outdoors; however in the same way
than outdoors two types of parameters are proposed:
The effect of these two groups of parameters is linked, because contaminants need a
given level of humidity to act on the corrosion process. A combination of contaminants could
have different effects than the sum of the individual effects. In ISO 11844-1 [29] it is
explained that the impact of temperature and relative humidity can not be expressed
72 F. Corvo, T. Pérez, Y. Martin et al.
according to ISO-9223. It could be explained taking into account that the nature of TOW is
very different to outdoors conditions.
Recent reports [30-31] on the use of atmospheric corrosion sensors based on changes in
electrical resistance showed that when there were no contaminants[29] , in tests of 100-110
h., corrosion rate was zero or insignificant. These sensors can determine changes in metal
thickness lower than one nanometer. However, in the presence of 0.08 ppm of SO2 or 20
μg/cm2 of NaCl in the system, changes in thickness where always detected over 75% of
relative humidity. Corrosion rate was determined at temperatures of 20, 30 and 40oC and the
Arrhenius equation was used to calculate the activation energy of the reactions. This method
is very similar to the natural conditions.
TOW depends mainly on the meteorological parameters, the nature of the metal, the
properties of the metallic surface and of the corrosion products layer formed. The most
important climatic factors on the corrosion process are relative humidity, sunshine hours,
temperature of the air and the metal surface, wind velocity and duration and frequency of the
rain, dew and fog.
Dew or condensation of humidity is considered an important cause of the corrosion of
metals. Its formation depends on the relative humidity and on the changes of temperature.
Dew does not wash the metallic surface so the concentration of pollutants is relatively high
and could be more aggressive than rain. Rain gives rise to the formation of a thick layer of
water and also adds corrosive agents such as H+ and SO42-;however it can wash away the
contaminants as well. It will depend on the intensity and duration of the rainfall.
The results of a wide evaluation carried out on different exposure conditions in Cuba
have already been published [3-4].The following Cuban corrosion test stations and conditions
were considered:
− Coastal test station under outdoor, sheltered and ventilated shed conditions
− Urban-industrial test station under outdoor and sheltered conditions
− Rural test station under outdoor, sheltered, ventilated shed and closed space
conditions
As was expected, the higher corrosivity corresponds to the coastal station. It was also
noted that the higher the corrosivity of the atmosphere, the higher the difference between the
outdoor corrosion rate and the indoor corrosion rate.
The atmospheric corrosion rate of metals depends mainly on TOW and pollutants;
however, if the differences in the corrosion process between outdoor and indoor conditions
are taken into account, the influence of direct precipitation such as rain is very important for
outdoor and negligible for indoor conditions. The acceleration effect of pollutants could
change depending on wetness conditions of the surface, so the influence of the rain time and
quantity should be very important in determining changes in corrosion rate.
A model was proposed which consider the influence of the relationship between rain
quantity/time and the interaction between pollutants at different TOW. TOW-ISO was
divided into two parts, when air temperature was lower than 25oC and when air temperature
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 73
was higher than 25oC (up to 35oC). In this way three different times of wetness are
considered: time due to rain (included independently of TOW-ISO), time including rain, dew
and fog (TOW-ISO at air temperature <25oC), and time when evaporation of the electrolyte
layer prevails (air temperature >25oC)-in some cases this last variable should also include
rarely time of rain. This model was developed based on previous results indicating that when
air temperature is over 25oC a significant diminution of the surface water layer takes place,
possible due to a significant increase in evaporation of this water layer.
By considering that the average corrosion rate is influenced principally by the deposition
rates of chloride and SO2 and adding the effect of cleansing of the metallic surface caused by
rain the following empirical model was proposed:
-
K = a + b [Cl-] TOW-ISO5-25 + c [Cl ] TOW-ISO25-35 + d [SO2]TOW-ISO5-25 + [SO2]
-
TOW-ISO25-35 + f [Cl ] Train + g [SO2] T rain + h mm/Train
where
K = Weight loss in g/m2 for the exposure period.
-
[Cl ] = deposition rate of chloride ions in mg/m2d
[SO2] = deposition rate of sulfur compounds in mg/m2d
TOW-ISO5-25 = Time of wetness estimated according to ISO-9223 for air temperatures
between 5 and 25oC. Lower temperatures are not considered because they are almost
nonexistent in Cuban climate.
TOW-ISO25-35 = Time of wetness estimated according to ISO-9223 for air temperatures
between 25 and 35oC.
Train = Time of rainfall (h) for the exposure period. This time is also included in TOW-
ISO, but its effect is significantly different.
mm = Amount of rain in mm for the period of exposure.
Data of corrosion rate of carbon steel, copper, zinc and aluminum together with different
TOW and contaminants were statistically processed (stepwise regression) and the following
results were obtained:
The variable [SO2] T rain in all cases was not statistically significant.
- -
K = 17.74 + [2.47 [Cl ] + 0.071 [SO2] ] TOW-ISO5-25 – 1.5 [Cl ] TOW-ISO25-35
r = 0.99 r2 = 99.0
Indoor conditions:
K = 31.55 + [2.02 [Cl-] + 0.13 [SO2]] TOW-ISO5-25 – 1.06 [Cl-] TOW-ISO 25-35
R = 0.99 r2 = 96.0
K = 21.29 + [0.62 Train + 1.59 TOW-ISO 5-25 – 1.11 TOW-ISO 25-35] [Cl-] + 0.09 [SO2]
TOW-ISO25-35
r = 0.98 r2 = 95.0
Metal: Copper
Outdoor conditions:
Indoor conditions:
K = 0.88 [Cl-] TOW-ISO5-25 + 0.17 [SO2] TOW-ISO 5-25
r 0.99 r2 = 97.0
Outdoor and Indoor conditions:
K = [1.03 [Cl-] + 0.24 [SO2] ]TOW-ISO 5-25 – 0.46 [Cl-] T rain + 0.19 mm/Train
r = 0.98 r2 = 94.0
Metal: Zinc
Outdoor conditions:
K = 12.22 + 0.98 [Cl-] Train
r = 0.98 r2 = 96.0
Indoor:
K = [1.98 [Cl-] + 0.13 [SO2] ]TOW-ISO5-25 – 1.09 [Cl-] TOW-ISO 25-35
r 0.99 r2 = 99.0
Outdoor and indoor:
K = 6.28 + [1.86TOW-ISO 5-25 – 0.96 TOW-ISO25-35] + 0.12 [Cl-] Train + 0.06 mm/Train
r = 0.98 r2 = 95.0
Metal: Aluminum
Outdoor conditions:
K = 0.46 + 0.96 [Cl-] Train
r = 0.96 r2 = 90.0
Indoor:
K = 0.80 [Cl-] TOW-ISO5-25 + 0.37 [SO2] TOW-ISO5-25
r 0.99 r2 = 96.0
Outdoor and indoor:
K = 0.55 + [1.81[Cl-] + 0.18 [SO2] ]TOW-ISO 5-25 –[ 0.24 Train + 0.89 TOW-ISO25-35]
-
[Cl ] - 0.22 mm/Train
r = 0.96 r2 = 89.0
The suggestion of dividing the time of wetness into three different contribution parts was
made in order to get a more quantitative approach step to study the atmospheric corrosion
process. The influence of time and quantity of rain is very important for characterizing
differences between indoor and outdoor corrosion. It can be observed that in all cases the
variable TOW 25-35 is affected by a negative sign, indicating a diminishing of corrosion rate
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 75
when this variables increases, what is according to the fact that when temperature is over
25oC the electrolyte layer significantly diminishes instead of existing a high relative humidity.
The cleaning effect of rain is important in the corrosion process. In many cases this
variable is significant, as well as the inclusion of time of rain as an independent variable in
explaining the influence of different parameters in atmospheric corrosion rate of basic metals.
It can be noted that the lower TOW is obtained for the storehouse having air conditioner
and the higher for the storehouse located in the coastal site and the metallic box located in the
rural site. According to these results, the higher possibilities of corrosion should be in these
last conditions if we only consider the definition of TOW and the values obtained. However,
it is very well known that the higher corrosion rate is usually obtained in outdoor or sheltered
conditions. This figure shows that the concept of TOW established on ISO 9223 could not be
used simply and many considerations have to be made. It should be limited to outdoor
conditions and some considerations have to be made respecting the nature and characteristics
of different meteorological parameters such as rain, dew, fog, i.e..
In other words, the concept of time of wetness (established in ISO 9223) shows
limitations, because the nature and the changes occurred in the electrolyte formed on the
metallic surface have clear differences between exposure conditions. There is a significant
difference between outdoor time of wetness and sheltered or indoor TOW.
60
50
Air conditioned storehouse
40
)
Average T
(
Ave.min.T
30 Ave.max T
p
Min T
Max T
20
10 no windows storehouse
0
Rural Coastal coastal Urban Rural Coastal
Outdoor and Heat trap
sheltered conditions
Figure 3. Changes in average, average maximum and minimum, maximum and minimum air
temperature depending on exposure condition and type of atmosphere for test stations located in the
western side of the Cuban Isle.
Actually, the general concept of TOW (time during which corrosion reaction occurs) is,
undoubtedly, correct and generally applicable, taking int account the electrochemical nature
of the atmospheric corrosion process, but the estimation made by ISO 9223 has to be limited
to specific conditions, mainly outdoors and taking into consideration the differences in the
corrosion reaction as a function of the way the water is present on the surface, because, for
example, rain forms an electrolyte in the metallic surface different to dew, and the formation
of the electrolyte during rain also causes a diminishing rate of deposition of contaminants in
the surface depending on the extent of rain, How can this complicated process be
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 77
represented?. Obviously, not by the simple concept established by ISO 9223. It should be also
considered the nature of the metal and the corrosion products.
100
90 no windows st orehouse
80
70
opened w indow s
60 storehouse
Average RH
Ave.min.RH
50 Ave.max.RH
Max RH
air condit ioned st orehouse
40 Min RH
30
20
10
0
Rural Coast al coast al Urban Rural Coast al
Out door and Heat t rap
shelt ered condit ions
Figure 4. Changes in average, average maximum and minimum, maximum and minimum air relative
humidity depending on exposure condition and type of atmosphere for test stations located in the
western side of the Cuban Isle.
Average annual corrosion rate of steel reported for test stations depending on exposure
conditions and type of exposure (see figure 6) is significantly different to the behavior of time
of wetness for the same conditions. It is perfectly explained based on the influence of
contaminants. Chloride deposition rate causes a significant acceleration of corrosion rate,
which is the reason why the maximum annual corrosion lost is reported for outdoor
conditions in the coastal zone, following sheltered and ventilated shed in the same coastal
zone. Urban industrial atmospheres (outdoor, sheltered and ventilated shed) are the second
one in weight lost and rural stations the less aggressive atmosphere. Regarding indoor
corrosion, the lower weight lost is determined in the storehouses and heat trap conditions
increase respecting the last one.
78 F. Corvo, T. Pérez, Y. Martin et al.
8000
No direct wat er precipit at ions
Closed st orehouse
7000
Rain, dew, conden-
sat ion, spray
6000
5000
3000
1000
0
Rur a l Rur a l Ur ba n Ur ba n
Out door a nd Ve nt i l a t e d she ds
she l t e r e d
Figure 5. Time of wetness at different exposure conditions and types of atmospheres calculated
according to ISO 9223 for test stations of the western side of the Cuban Isle.
4000
3500
Corrosion rate (g/m2a)
3000
2500
Rural
2000 Urban-industrial
Coastal
1500
1000
500
0
Outdoor Sheltered Ventilated sheds Storehouses Heat trap
conditions
Type of exposure
Figure 6.Average corrosion rate (g/m2a) depending on exposure condition and type of atmosphere for
test stations located in the western part of the Cuban Isle
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 79
As it can be observed, in the storehouse where there is determined the maximum time of
wetness, it is reported the lower corrosion rate. It is explained based in the fact that in this
case TOW consists mainly in air humidity without the contribution of any precipitation or
condensation (no rain, no dew, no significant condensation). It is a demonstration that the
estimation of time of wetness according to ISO 9223 presents limitations.
Changes in steel corrosion rate for different exposure conditions and types of atmosphere
are better explained based on average deposition rate of Chlorides, as can be observed on
figure 5. It can be note that Chloride deposition is higher in coastal stations, as it should be
expected, and it is in agreement with a higher weight lost for these stations.
The lower values of Chloride deposition are reported for the storehouses, where
coincidently, there are reported the lower weight losses. Under heat trap conditions there is a
given increase in Chloride deposition, that is why corrosion rate of steel is higher than in
storehouses.
1000
100
Deposition rate (mg/m2d)
Rural
10 Urban-industrial
Coastal
0,1
Outdoor V.Shed Storehouse Heat-trap
Type of exposure
Figure 7. Annual average Chloride deposition rate for test station located at the western side of the
Cuban Isle depending on exposure conditions and type of atmosphere.
On figure 7 there is represented Average Chloride Deposition rate determined in Cuba for
different types of atmosphere and exposure conditions. The results are very similar to those
presented on figure 6, particularly respecting outdoor and ventilated shed conditions. Chloride
deposition is determined under shelter, that is why it is assumed the same value for ourdoor
and sheltered conditions; however, it is very well known, the significant influence of
precipitations upon outdoor corrosion and its negligible effect under sheltered conditions. In
80 F. Corvo, T. Pérez, Y. Martin et al.
storehouses and heat trap conditions the influence of Chloride deposition significantly
diminishes; however, under heat trap conditions the higher corrosion of steel is reported for
the coastal test station.
50
45
40
35
Deposition rate (mg/m2d)
30
Rural
25 Urban-industrial
Coastal
20
15
10
0
Outdoor V. Shed Storehouse Heat-trap
Type of exposure
Figure 8. Annual average sulfur compound deposition rate for test stations located in the western side of
the Isle of Cuba depending on exposure conditions and type of atmosphere.
The average deposition of sulfur compounds found at the Cuban western side test station
is presented on figure 8. As can be expected, the higher sulfur compound deposition rate
corresponded to the urban industrial station in outdoor conditions. There is no report for
ventilated shed. Part of the sulfur compound deposition for coastal stations should not be
attributed to gaseous SO2 or SO3, because it is well known that airborne salinity contains
sulphate ions and the procedure used for determining sulfur compounds deposition is sensible
to sulphate ions (see ISO 9225 [31]). A direct relationship between average corrosion rate and
sulfur compound deposition is not observed as in the case of Chloride deposition. It confirms
the importance of Chloride deposition in determining corrosion aggressivity in Cuba.
Chloride ion is one of the most important natural pollutants influencing corrosion;
particularly in the tropical humid conditions of Cuba and Campeche in Mexico; however, its
role may change depending on climate. A very humid climate can cause a fast leaching of the
chloride ions and decreases its effect on the acceleration of corrosion rate. A different
acceleration rate has been reported for chloride ions between Eastern and Western Caribbean
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 81
[2]. In a given climate the influence of rain could change the acceleration caused by chloride
ion on metal corrosion. Differences between rainy and dry periods could be important in
determining such acceleration. The electrochemical mechanism of corrosion by chloride ions
does not change, but the time during which chloride ions acts on the metallic surface and their
concentration may appreciable change, depending on factors such as climatic conditions and
nature of corrosion products.
A research about the role of the rainfall characteristics on the acceleration rate caused by
chloride ions between two stations having noticeable differences in rain regime was already
published[32]. The influence of chloride ions at Medellin and Havana corrosion stations was
studied by submitting samples of steel and copper to salt spray during exposure to the open
atmosphere. Up to 12 months of exposure, the acceleration on the deterioration caused by
chloride ions was notably up to 12 months of exposure, being higher at Havana station for
steel and copper.
The corrosivity category for steel and copper (C3) in natural conditions in Havana
increases to C5 for copper and over C5 for steel when salt spray is applied; however at
Medellin, copper corrosion increases to C4 for copper and C5 for steel under the application
of salt spray.
The acceleration rate caused by the addition of a salt spray under natural conditions
causes a higher acceleration of corrosion rate at Havana. For copper, the difference does not
appreciably change with time; but it does for steel. A more aggressive action of chloride ions
is observed in the case of steel.
The remarkable difference in the acceleration caused by chloride deposition rate in
Havana and Medellin could be due to the considerable difference in the rain regime between
both sites. Other characteristic of the environment could also have influence, but rain should
play an important role.
Samples were submitted to the same artificial chloride deposition rate at Havana and
Medellin stations (supposing that differences concerning natural values of chloride deposition
rate are not very significant). Samples were prepared from the same origin material, so the
differences on the corrosion behavior can be associated only to climatic characteristics. In the
case of Medellin the washing effect of rain should be higher than in Havana because the rain
amount is more than four times than the reported for Havana during the exposure period. It
explains why a lower acceleration rate of chloride ions is obtained at the former station.
Based on this fact, a model for studying the influence of the washing or cleaning effect of rain
in the determination of the acceleration rate of chloride ions is proposed.
The bi-logarithm equation for atmospheric corrosion establishes that:
K=atb
K = a t b[Cl]c (W/D)d
82 F. Corvo, T. Pérez, Y. Martin et al.
where K = mass loss (g/m2); t = time of exposure (months); [Cl] = chloride deposition
rate (mg/m2 d); w = rainfall amount (mm); D = number of rainy days.
A regression analysis including copper mass loss data for Havana and Medellin
showed the following results:
A good data fit is also obtained for copper. It confirms that the complex chloride
deposition rate–rain regime is important for determining mass loss of copper and steel. In this
model TOW-ISO has not been used and a good fitness has been obtained.
It confirms that the acceleration rate caused by chloride ions on atmospheric corrosion of
steel and copper depends on the characteristics of rain regime. For a place having high
amount and time of rain, a lower acceleration on corrosion rate should be expected for a given
chloride deposition rate
The washing and cleaning effect of rain is not included and is far from the concept of
TOW established on ISO 9223 standard.
where:
[Cl-] W.C = Chloride deposition rate determined using Wet Candle method
[Cl-] D.P = Chloride deposition rate determined using Dry Plate method
Chloride deposition rate was determined using Wet Candle and Dry Plate methods in the
corrosion stations Santiago de las Vegas (rural-urban), Casa Blanca (industrial-marine-urban),
Via Blanca (industrial-urban-marine) and Cojimar (marine).
On figure 9 it can be noted a significant difference between Steel corrosion rate and
Chloride deposition between the north and the south shores. It is caused because, in general,
trade winds in the north shore come from the Ocean and in the south shore from the earth. In
addition, cold fronts always come from the north. The territory is flat, presenting only small
hills, so Chloride deposition reaches almost all the territory. It can be seen that even in places
located at 15-30 km from the north seashore a significant Chloride deposition rate is
determined (3.4-8.5 mg/m2d-classified as S1 according to ISO-9223). Sulfur compounds
deposition has also some relation with Chloride deposition. The higher values correspond to
coastal and industrial stations. It means that, taking into account that the determination using
alkaline surfaces is sensible to different sulfur compounds, it includes, in addition to sulfur
oxides, sulfates coming in airborne salinity. Another possible sulfur compound determined
could be H2S
4000 600
3500
500
2000 300
1500
200
1000
100
500
0 0
0.01 0.15 2.0 2 2.0 3.0 3 6 15 18.0 30 30 30.0 48.0 South shore
North shore Distance to the north sea shore (km)
Figure 9. Changes in annual Corrosion rate of mild steel flat samples, annual average Chloride
deposition rate determined by the dry plate method and annual average Sulfur compounds deposition
rate determined by alkaline surfaces method in the west side of the Cuban Isle.
Changes in TOW-ISO are negligible between all the test stations of the Cuban western
side. All values are classified as τ4 according to ISO 9223. In these conditions main changes
depend on contaminants deposition. Due to the hygroscopic properties of the salts deposited
84 F. Corvo, T. Pérez, Y. Martin et al.
in the surface there is no doubt that corrosion should occur at a relative humidity lower than
80%, so in this case the use of TOW-ISO has no sense.
Deposition rate of Chloride at Campeche, Veracruz, Puerto Progreso and Puerto Morelos
500
450
Campeche
Chloride deposition (mg/m2d)
400 Veracruz
Puerto Morelos - Puerto Progreso
350
300
250
200
150
100
50
0
-10 0 10 20 30 40 50 60 70 80
distance to the shoreline (km)
Figure 10. Annual average of deposition rate of Chlorides for different sites of the Yucatán Peninsula:
Puerto Progreso and Puerto Morelos (Yucatán and Quintana Roo States), Campeche and Veracruz
States.
Steel corrosion rate at Campeche, Veracruz, Puerto Progreso and Puerto Morelos
(flat specimens)
3500
3000 Campeche
Veracruz
Puerto Morelos - Puerto Progreso
Corrosion rate (g/m )
2500
2
2000
1500
1000
500
0
-10 0 10 20 30 40 50 60 70 80
distance to the shoreline (km)
Figure 11. Annual average corrosion rate of carbon steel at Campeche, Veracruz and Puerto Morelos as
function of distance to the shoreline.
On figure 10 the annual average Chloride deposition rate determined at three zones of the
Yucatán Peninsula is presented: Puerto Progreso and Puerto Morelos in the mouth of the Gulf
of Mexico and Campeche and Veracruz inside this Gulf as function of the distance to the
shoreline. As can be seen, Chloride deposition is very similar at Puerto Progreso, Puerto
Morelos and Veracruz; however, the values of Chloride deposition are significantly lower for
Campeche. It could be explained based on the fact that predominant winds in Campeche most
of the year are from earth to the Gulf of Mexico. Another factor is that sea Campeche usually
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate 85
has no wave movement. It is a similar situation to the south shore of the Cuban Isle where a
lower Chloride deposition is determined.
Changes of corrosion rate of steel as function of distance to the shoreline are lower
respecting Veracruz. It is in perfect agreement with the influence of Chloride deposition. In
the case of Puerto Morelos and Puerto Progreso it is also reported a significantly higher
corrosion rate respecting Campeche.
Average Chloride
Average corrosion Standard Standard TOW-ISO
Climatic season deposition rate
rate of steel (g/m2) deviation deviation (h)
(mg/m2d)
Summer(May/oOct) 569.4 167.6 225.0 107.7 2112
Winter(Nov/Apr) 2111.5 892.8 486.3 174.2 2000
Summer/Winter 2.53 -- 4.34 -- 1.06
Table VIII. Average Chloride deposition rate for differents test stations of Campeche
Chloride ions move through the corrosion products layer and arrive at metal surfaces
producing a notable acceleration in corrosion rate. A catalyser role has been suggested for
anions in the corrosion process. Chloride ions are very soluble, that is why they are easily
removed from metallic surfaces due to liquid precipitation. Usually, a low concentration of
chloride ions is determined in corrosion products on outdoor exposures. A relation between
chloride concentration present in corrosion products and corrosion rate has been reported.
Rust samples scraped off from exposed AISI 1019 steel surfaces were ground to fine size
in a morter with a pestle. A fraction of the samples was subjected to a salts extraction process
using distilled water to determine the concentration of most common ions by chemical
analysis and to study changes in adsorption isotherm after elimination of hygroscopic salts.
Samples were obtained from AISI 1019 steel coupons, exposed for up to six months at coastal
and rural locations.
The water adsorption isotherms were determined by the liquid-volumetric method[21]
The rust samples were put in an atmosphere with certain amount of water up to obtaining an
established equilibrium, the pressure is measured and the water adsorbed may be calculated.
The experimental isotherms were fitted to the BET and Langmuir isotherms, selecting the
best fit. The values of the monolayer capacity, the specific surface and the net heats of
adsorption were calculated.
The adsorption isotherms for metallic surfaces are reported in the literature; however, an
important part of the atmospheric corrosion process takes place under rust layers, which play
a decisive role in the long-term course of corrosion because of its sorption capacity for water.
The influence of the chloride and sulfate anions has a real effect only when the corrosion
products layer is already formed. Thus, the adsorption isotherms of the steel corrosion
products formed in different atmospheres were determined.
It can be seen that chloride, sulfate and bicarbonate concentrations are higher in the
sample from Viriato station, because here the atmosphere is more contaminated.
Nevertheless, the relation between the chloride concentration in the corrosion products from
Viriato and Quivican is (24.64/15 = 1.65) far smaller than the Cl-DR relation (495.68/3.9 =
127.1) obtained in these two atmospheres. This is because the corrosion products are washed
away from the corroding surface during rain and chlorides are easily eliminated from the
metallic surface. It confirms the role of rain in washing and cleaning contaminants in the
corrosion products layer. Chloride collectors are protected from liquid precipitations. There
should be periods of very high chloride concentration with rapids changes due to liquid
precipitations. In the case of sulfate, the ratio Viriato/Quivican is very similar to the
atmosphere (17.28/3.84 =4.5 and 34.55/10.5 = 3.3, respectively). It seems that liquid
precipitations do not appreciably affect sulfate concentration in corrosion products.
The value of steel corrosion rate ratio for Viriato and Quivican stations at six months of
exposure (1154.67/108.37 = 10.65) is between Cl-DR and chloride concentration in corrosion
products ratio and is higher than sulfate. There should be rapid changes in chloride
concentrations due to liquid precipitations and arrival of aerosol particles. Changes in sulfate
concentration should be lower than for chloride. The corrosion rate in coastal zones should
depend mainly on a sum of time of wetness at different chloride and sulfate concentrations.
88 F. Corvo, T. Pérez, Y. Martin et al.
Bicarbonate ions concentration at Viriato is higher as the value of pH increases. Thus, the
corrosion process and the corrosion products transformation occur to a higher pH and this can
lead to changes in the corrosion mechanism and in the nature of corrosion products formed.
Sodium and potassium concentration (Na+ should be the main constituent) is higher at
Viriato according to a very high anionic concentration and also a high conductivity is
determined.
Water adsorption should be caused by salts content but, perhaps, the presence of high
levels of contaminants can determine the formation of a corrosion product having a more
“hygroscopic” nature. To make clear this possibility, adsorption isotherm of corrosion
products from Viriato corrosion station after elimination of salts were determined . The values
of adsorption are lower after salt elimination, but higher than those corresponding to
Quivican. It could be due to a more “hygroscopic” nature of corrosion products formed in
coastal stations.
The superficial characteristics of atmospheric corrosion products of steel depend on the
type of atmosphere where the sample has been exposed. The way of adsorption of the
corrosion products obtained in the coastal atmosphere is polymolecular due to a higher
content of salts. This makes easier the presence of water in the metal-corrosion products
interface and determines a high corrosion rate. The adsorption of water of a corrosion product
formed in a rural zone obeys a Langmuir isotherm, i.e. a monomolecular adsorption takes
place. It causes a lower corrosion rate.
Chloride ions are easily washed away from the metallic surface by precipitation, while
sulfate ions concentration remains more constant. It means that corrosion takes place under
very variable concentrations of chloride ions and in the presence of less variable
concentration of sulfate ions. Changes in the amount of chloride ions concentration should
depend on the particular rain regime of the place. A notable difference has been found
between chloride deposition rate determined by collectors protected from liquid precipitations
and chloride content in corrosion products.
The determination of adsorption isotherms offers the possibility to study the atmospheric
corrosion process when the metal is covered by a corrosion products layer.
Conclusion
TOW considered as the time during which corrosion occurs is an important parameter in
atmospheric corrosion of metals. It defines the possibility for atmospheric corrosion to occur
based on its electrochemical nature, but corrosion rate will depend mainly in the acceleration
caused by contaminants deposition and other factors.
Acknowledgments
The authors appreciate the contribution of M. Sc. Miguel Ramón Sosa Baz (Corrosion
Research Center, University of Campeche, Campeche, Mexico) and Tec. Eva González
Mellor and Julia Pérez Acosta (National Center for Scientific Research (CNIC), Havana City,
Cuba)
References
[1] F. Corvo, N. Betancourt, A. Mendoza, Corrosion Science, vol. 37, No. 12, p. 1889-
1901, 1995.
[2] F. Corvo, C. Haces, N. Betancourt, L. Maldonado, L. Veleva, M. Echevarria, O.T. de
Rincón, A. Rincón, Corrosion Science, vol. 39, No. 5, p. 823-833, 1997.
[3] R. Mendoza, F. Corvo, Corrosion Science, vol. 41, 1999, p.75-86.
[4] A.R. Mendoza, F. Corvo, Corrosion Science, vol. 42, 2000, p. 1123-1147
[5] L. Veleva, L. Maldonado, British Corrosion Journal, vol. 33, No.1, p. 53-57,1998.
[6] D. C. Cook, A.C. Van Orden, J. Reyes, S.J.Oh, R. Balasubramanian, J.J. Carpio, H.E.
Townsend, Atmospheric Corrosion in Marine Environments along the Gulf of
México, Marine Corrosion in Tropical Environments, ASTM STP 1399, S.W. Dean,
G. Hernández Duque-Delgadillo and J. Bushman, Eds. American Society for Testing
and Materials, West Conshohocken, PA, 2000.
90 F. Corvo, T. Pérez, Y. Martin et al.
[7] M. Sosa, A. del Sol, Revista de Climatología, Cuba, vol. 2, 2004, pp. 1-4.
[8] L. Mariaca, J. Genesca, J. Uruchurtu, L. Salvador. Corrosividad Atmosférica
(MICAT-México). Programa Iberoamericano de Ciencia y Tecnología para el
Desarrollo (CYTED), 1999. Editorial Plaza y Valdés S.A. de C.V., México.
[9] ISO 9223:1992. Corrosion of metals and alloys. Corrosivity of atmospheres.
Classification.
[10] Leygraf, T.E. Graedel, Electrochemical Society Series, Wiley Intersciences, 2000
[11] ISO/WD 9223. Corrosion of metals and alloys. Corrosivity of atmospheres.
Classification
[12] R. Pascual, F. Corvo, Rev. Iberoamericana de Corrosión y Protección, mayo-junio,
1980.
[13] F. Corvo, J. Rocha. Characterization of environments and their influences on Copper
at different atmospheres and exposure conditions in tropical climate, International
Corrosion Congress (ICC), Granada, Spain, 2002.
[14] L. Veleva, M. A. Alpuches-Aviles, Outdoor atmospheric corrosion, ASTM, STP
1421, H.E. Townsend, Ed. American Society for Testing and Materials International,
West Conshochoken, PA, 2002.
[15] Tooru Tsuru, Ken-Ichiro Tamiya, Atsushi Nishikata, Electrochimica Acta 49 (2004)
2709–2715.
[16] Li Bian, Yongji Weng , Xiangyi Li, Electrochemistry Communications, 7 (2005),
1033-1038.
[17] Jibiao Zhang, Jia Wang, Yanhua Wang, Electrochemistry Communications, 7 (2005)
443–448.
[18] Wang Jia, Zhang Jibiao. The effect of micro-droplets formation caused by the
deliquescence of the deposited salt particle on atmospheric corrosion of metals.
Proceedings16th International Corrosion Congress, Beijing, China, September 19-24,
2005.
[19] J. Tidblad , A. A. Mikhailov and V. Kucera, “Application of a Model for prediction of
atmospheric corrosion in tropical environments”, Marine Corrosion in Tropical
Environments, ASTM STP 1399, S.W. Dean, G. Hernández Duque-Delgadillo, and J.
B. Bushman, Eds., American Society for Testing and Materials, West Conshohocken,
PA, 2000.
[20] Iu. N. Mikhailovskii, Atmospheric corrosion of metals and protection methods (in
russian). Editorial “Metallurgiia” , Moscow, 1989.
[21] F. Corvo, A. R. Mendoza, M. Autie, N. Betancourt, Corrosion Science, vol. 39, No. 4,
pp. 815-820, 1997.
[22] José Rocha Andrade da Silva Editor. Productos electro-electrónicos en ambientes
tropicales, Campinas, Sao Paulo, Sitta Gráfica 2003.
[23] Leygraf. Indoor atmospheric corrosion. Proceedings 15th International Corrosion
Congreso, Granada, Spain, September 22-27, 2002.
[24] F. Corvo, A. D. Torrens, N. Betancourt, J. Perez, E. Gonzalez, Corrosion Science, vol.
49, No. 2, February 2007, , pp. 418-435.
[25] S. Cole, R. Holgate, P. Kao and W. Ganther, Corrosion Science, vol. 37, No. 3, 1995,
pp. 455-465.
[26] M. Stratmann, H. Streckel, Corrosion Science, vol. 30, No. 6/7, 1990, pp. 697-714.
Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate… 91
[27] M. Stratmann, H. Streckel, K. T. Tim, S. Crockett, Corrosion Science, vol. 30, No.
6/7, 1990, pp. 715-734.
[28] ISO 11844-3.-Corrosion of metals and alloys. Classification of corrosivity of indoor
atmospheres. Measurement of environmental parameters affecting indoor corrosivity.
[29] ISO 11844-1.-Corrosion of metals and alloys. Classification of corrosivity of indoor
atmospheres. Determination and estimation of indoor corrosivity.
[30] J.P. Cai, S.B. Lyon, Corrosion Science, vol. 47, (2005) 2956-2973.
[31] Seon Yeob Li, Young-Geun Kim, Sungwon Jung, Hong-Seok Song, Seong-Min Lee,
Sensors and Actuators B, vol. 120, 10 January 2007, pp. 368-377.
[32] F. Corvo, J. Minotas, J. Delgado, C. Arroyave, Corrosion Science, vol. 47, 2005
p. 883-892.
[33] ISO-9225: 1992. Corrosion of metals and alloys. Corrosivity of atmospheres.
Measurement of pollution.
[34] F. Corvo, R. Marbot, Revista CENIC (Ciencias Químicas), vol. 15, No.1, 1984
In: Electroanalytical Chemistry: New Research ISBN: 978-1-60456-347-4
Editor: G. M. Smithe © 2008 Nova Science Publishers, Inc.
Chapter 3
Thomas Z. Fahidy
Department of Chemical Engineering, University of Waterloo, Waterloo
Ontario, Canada, N2L 3G1
Abstract
Randomness, independence and trend (upward, or downward) are fundamental concepts in a
statistical analysis of observations. Distribution-free observations, or observations with
unknown probability distributions, require specific nonparametric techniques, such as tests
based on Spearman’s D – type statistics (i.e. D, D*, D**, Dk ) whose application to various
electrochemical data sets is herein described. The numerical illustrations include surface
phenomena, technology, production time-horizons, corrosion inhibition and standard cell
characteristics. The subject matter also demonstrates cross fertilization of two major
disciplines.
List of Symbols
A. ordinal form of an observation set; A* same with rank ties
B ordinal form of an observation test (to be tested for association/independence vis-à-vis set
A); B* same with rank ties
b total number of blocks in two-factor analysis
D Spearman’s D-statistic, defined by Eq.(2); D* the same with rank ties, defined by Eq.(8)
D** D-statistic extended to ordinal forms of two observation sets, defined by Eq.(11)
d numerical value of D
Ecorr corrosion potential
94 Thomas Z. Fahidy
I. Introduction
Nonparametric statistics (NPS) differs primarily from its traditional, distribution-based
counterpart by dealing with data of unknown probability distributions. Its principal
attractiveness lies, in fact, in not requiring the knowledge of a probability distribution. NPS is
especially inviting when the assumption of normal distribution of small data sets is hazardous
(if at all admissible), even if NPS-based calculations are more time consuming than in
traditional statistics. The steadily growing importance of NPS has been amply demonstrated
by numerous textbooks and monographs published within the last few decades, e.g. [1-7].
Exploration of the scope of NPS in electrochemical science and engineering has so far
been rather limited. The estimation of confidence intervals of population mean and median,
permutation-based approaches and elementary explorations of trends and association
involving metal deposition, corrosion inhibition, transition time in electrolytic metal
deposition processes, current efficiency, etc.[8] provides a general framework for basic
applications. Two-by-two contingency tables [9], and the analysis of variance via the NPS
approach [10] illustrate two specific areas of potential interest to electrochemical process
analysts.
The current subject matter deals with randomness, independence and trend concerning
small sets of electrochemical observations or measurements, whose probability distribution is
unknown, and the assumption of an even approximately normal distribution would be
statistically unsound. Under such circumstances the theoretical null-distribution related to the
hypothesis H of lack of randomness, independence and lack of trend, has to be established
from the data themselves on the basis of equal probability of all possible data arrangements.
The Application of D-Statistics Based Tests of Randomness, Independence… 95
The first illustration is fashioned after Problem 29 [13] of iron corrosion in deoxygenated
aqueous acid. It is assumed that four identically shaped iron-containing specimens labeled A,
B, C, D are chosen randomly, and inserted in the acid carrying a corrosion inhibitor. The
percentage reduction in the specific rate loss is 93.3(A), 97.3(B), 96.7(C) and 90.0(D). As
shown in Table 1, the hypothesis H of no treatment effect (i.e. the hypothesis of all rankings
being equally possible) can be stated as
Table 1. Establishment of the null distribution of ranks and the D-statistic in testing the
hypothesis of upward trend in the rate-loss reduction data in Section II.
i 1 2 3 4
Ti 3 1 2 4
2 3 1 4
1 3 2 2
1 4 2 3
Illustration for d = 6
In all four cases the sum of the squared differences (Ti – i)2 is 6. Hence, P[D = 6] = 4(1/4!) = 1/6
1 1
PH (T1 = t1 ; T2 = t 2 ;..., TN = t N ) = = (1)
N ! 4!
in terms of the probability that the ranks Ti take on any particular set of values ti ,i = 1,…,N,
respectively. This is also the desired null distribution of the ranks, expressing formally the
hypothesis of randomness. The alternative (or counter) hypothesis HA of upward trend can be
tested via the numerical value of the Spearman D-statistic
N
d = ∑ (t i − i ) 2 (2)
i =1
by determining the cumulative probability P[D ≤ d]. As shown in Table N [4,p.433], there
exists an approximately 79% probability that the hypothesis of randomness can be sustained.
The normal distribution-based approximation, correct in a strict sense for large observation
sets:
d − EH
P[ D ≤ d ; N ] ≅ FN ( ) (3)
VH
with expectation
N3 − N
EH ≡ = 10 (4)
6
and variance
( N − 1) N 2 ( N + 1) 2
VH ≡ ≈ 33.33 (5)
36
yields a somewhat lower probability of about 75%. If a continuity correction is applied, i.e. (d
+ 1/2) is used in Eq.(3) instead of d, the probability is about 78%.The alternative hypothesis
of upward trend cannot be supported.
The next illustration portrays the high level of cautiousness carried by the testing-against-
trend approach. The case in point is the production of caustic soda in seven major industrial
The Application of D-Statistics Based Tests of Randomness, Independence… 97
countries in the years 1970, 1975 and 1979 [14], with ranking distributions assembled in
Table 2. Since in the presence of ties, the modified D-statistic D* related to the Ti* ranks:
N
D * = ∑ (Ti* − i ) 2 (6)
i =1
can also have a fractional value {computed as d* = ∑(ti* - i)2 }, Table N [4, p.233] provides
probabilities for the occurrence of integer values of D* flanking the computed value. The
appropriately modified normal approximation with parameters
N3 − N 1 g
EH (D ) ≡
*
− ∑ (d 3j − d j ) (7)
6 12 j =1
and
g
( N − 1) N ( N + 1)
2 2 ∑ (dj =1
3
j −dj)
VH ( D * ) = [1 − ] (8)
36 N3 − N
provides the probability estimate via Eq.(3). The obvious upward trend in Case B is indicated
to be not significant at an about 17% probability of D = 0. At N = 4 the probability is
significant, whereas at N > 4, it is highly significant (0.0417 at N = 4; 0.0083 at N = 5; 0.0014
at N =6, etc.), with gradually decreasing significance probabilities. As shown in Section III.3,
a much more realistic result is obtained by appropriate two-factor analysis, employing
Friedman’s FM-statistic.
98 Thomas Z. Fahidy
P[ D ≤ ∑ Dk ] ≅ FN (
∑D k − E H ( ∑ Dk )
(9)
V H ( ∑ Dk )
with parameters
N3 − N
E H (∑ Dk ) ≡ k (10)
6
and
( N − 1) N 2 ( N + 1) 2
V ( ∑ Dk ) ≡ k (11)
36
is not significant. If the salt effect is a-priori assumed to be non-random, this case is an
obvious candidate for two-factor analysis. So is the similar case of the corrosion [16] of Al-
2.5Mg alloy in 0.5 mol dm-3 NaCl solution, in the presence of inhibitor sinapic acid(cinamic
acid, 3,5-dimethoxy-4-hydroxy-sinapine), portrayed in Table 4. The hypothesis of no upward
trend in corrosion potential (i.e. no significant rpm effect) cannot be rejected, if the column
entries are considered simply as random replicates. By contrast, upward trend in corrosion
current is highly significant. An identical result can be obtained for inhibition efficiency, not
discussed here.
The Application of D-Statistics Based Tests of Randomness, Independence… 99
Table 4. Dk – based analysis of the ranking distribution of the corrosion potential (Ecorr)
and corrosion current density (icorr) of an Al – 2.5Mg alloy. The brackets contain mV
and μA cm-2 values, respectively [16]
(a)Corrosion potential
Inhibitor conc’n mol dm-3 Rotation speed, rpm 1 0 2 150 3 2000 Dk (&)
0 1 (0.771) 3 (0.776) 2 (0.773) 2
1 1 (0.767) 2 (0.768) 3 (0.771) 0
10 2 (0.770) 1 (0.768) 3 (0.773) 2
50 1 (0.771) 2 (0.775) 3 (0.776) 0
100 3 (0.800) 1 (0.772) 2 (0.773) 6
500 3 (0.820) 2 (0.806) 1 (0.774) 8
(&) computation pattern as in Table 3
∑Dk = 18; EH(∑Dk) = 6(33-3)/6 = 24; VH(∑Dk) = 6[(2)(32)(42)/36] = 48
P[∑Dk ≤ 18] ≈ FN[(18.5-24)/√48] = FN(-0.79) = 0.2148
Inhibitor conc’n mol dm-3 Rotation speed, rpm 1 0 2 150 3 2000 Dk (&)
0 1 (1.44) 2 (2.28) 3 (2.84) 0
1 1 (0.86) 2 (1.63) 3 (2.32) 0
10 1 (0.72) 2 (1.40) 3 (2.10) 0
50 1 (0.62) 2 (1.22) 3 (1.93) 0
100 1 (0.51) 2 (1.02) 3 (1.68) 0
500 1 (0.38) 2 (0.79) 3 (1.40) 0
(&) computation pattern as in Table 3
∑Dk = 0; EH(∑Dk) = 24; VH(∑Dk) = 48; P[∑Dk = 0] ≈ FN[(0-24)/√48] = 2.7x10-4
The D-statistic concept can be readily extended to the two-factor case, where (Ai ; Bi) are
corresponding ranks, i.e. if A and B are the ordinal form of observation sets (the A,B notation
is used instead of Lehman’s (R,S) notation in order to avoid confusion between similar
symbols). If rank positions are tied, they are replaced by their mid-rank, yielding modified
distributions A* and B*. By a straightforward extension of Eq.(6), the modified D-statistic is
computed as [17]
N
D ** ≡ ∑ ( Ai* − Bi* ) 2 (12)
i =1
N3 − N 1 e 1 g
EH (D ) ≡
**
− ∑ (d j − d j ) − ∑ ( f j3 − f j )
3
(13)
6 12 j =1 12 j =1
100 Thomas Z. Fahidy
and
e g
( N − 1) N ( N + 1)2 2 ∑ (d
j =1
3
j −dj) ∑( f
j =1
j
3
− fj)
VH ( D ) ≡
**
[1 − ][1 − ] (14)
36 N3 − N N − N3
where dj and fj are the number of mid-ranks in set A and B in increasing order, their total
numbers being e and g, respectively.
The first electrochemical application of the D**-statistic deals with the lack- of-
association (i.e. independence) hypothesis concerning current efficiency and current load in
diaphragm-type industrial scale chlor-alkali cells [18]. Table 5 demonstrates that the two
factors are independent with the understanding that the current efficiency/current load
relationship may indirectly be influenced by other technical variables, e.g. cell potential, and
impurities.
Table 5. Testing the independence of current efficiency and current load via eight
different diaphragm-type chlor-alkali cells [18]
Table 6. Testing the independence of impedance parameters and potential sweep rate in
surface oxide formation [19]. The observed values are in brackets
Oxide
Electrolyte Surface/species Constant.phase/ Ox.film
Sweep rate filmResist
Resistance Parameter element thickness
(mV s-1) ance
(Ω cm2) (106Ω-1sncm-2) exponent (nm)
(kΩ cm2)
1 (10) 1 (26.4) 1 (5.36) 4 (0.95) 5 (320) 4.5 (1.98)
2 (20) 4.5 (26.9) 2 (5.37) 1.5 (0.94) 4 (316) 4.5 (1.98)
3 (30) 4.5 (26.9) 4 (5.40) 4 (0.95) 2.5 (315) 2.5 (1.97)
4 (40) 2 (26.5) 5 (5.41) 4 (0.95) 2.5 (315) 1 (1.96)
5 (50) 3 (26.8) 3 (5.39) 1.5 (0.94) 1 (311) 2.5 (1.97)
D** 16.5 6 22.5 38.5 34
EH(D**) 19.5 19.5 19.5 19.5 19.5
VH(D**) 95 100 75 95 90
z (#) - 0.308 - 1.30 0.346 1.949 1.581
FN(z) 0.3783 0.0968 0.6368 0.9744 0.9429
Table N 0.392-0.475 0.117 > 0.525 > 0.525 > 0.525
(#) z ≡ [(D** - EH(D**)]/√VH(D**)
Table 8.Rank distribution analysis of the SEC versus current density, cell potential, and
degree of efficiency in industrial alkaline bipolar water electrolyzers at 80 0C electrolyte
temperature and 1 bar pressure [21]. The bracketed numbers are observed values.
Table 8 depicts tests against trend involving current density, cell potential, degree of
efficiency, and specific energy consumption, the latter considered as the trend variable.
Critical D** data lead to conservative predictions, but the normal approximation indicates a (i)
strongly upward SEC/cell potential, and a strongly downward SEC/degree of efficiency trend,
and (ii) a not significant SEC/current density trend, although FN = 0.06 is only slightly larger
than the significant 0.05 value. On account of symmetry, the distributions in Table N [4,
p.433] are symmetric around (N3-N)/6, and FN(z) = 1 – FN(-z), hence the testing of downward
trend is essentially the “mirror image” of the upward trend test. Other entries in [21] have
been excluded due to their different operation protocols (cell type, pressure) in order to avoid
“dilution” due to too many (possibly interacting) variables.
III. Discussion
III.1 Comparison with Nonparametric Analysis of Variance
A close inspection of Table 4 raises the question whether inhibitor concentration should
be considered as a block factor, preferably requiring a nonparametric randomized complete
block experiment approach. Employing for the corrosion potential the Friedman statistic, e.g.
[22-24]:
k
12
FM = ∑
bk (k + 1) i =1
Ri2 − 3b(k + 1) (15)
The Application of D-Statistics Based Tests of Randomness, Independence… 103
in the absence of rank ties, and in view of b = 6, k = 3, R1 = R2 = 11, and R3 = 14, the
computed FM = 1 is well below even the 10% critical value 5.333 [25], as well as the
approximate critical χ20..5(2) = 1.386 (the Friedman statistic has an approximately chi-square
distribution). The blocking effect, therefore, is insignificant. As for the corrosion current,
since R1 = 6; R2 = 12; R3 = 18, the computed FM = 12 is highly significant, since it is also the
0.1% critical value; alternatively, χ2(2) = 12 falls between the 0.5% critical value 10.60 and
the 0.1% critical value 13.82. Similarly, in the case of the membrane permeation flux (Table
3), with b = 5, k = 2, R1 = 7, R2 = 8, FM = 0.2 being well below even the 25% critical χ2(1) =
1.32, the blocking effect is found to be insignificant.
Apparently the oldest nonparametric measure of the strength of association between two
factors [26], the Spearman rank correlation coefficient
6D
rs = 1 − (16)
N3 − N
yields rs = 1 when D = 0 (fully upward trend, regardless of the scatter of data!), and it
approaches rs = -1 for large values of D, i.e. when the downward trend is very strong.
Although rs may be regarded as an “…equivalent test statistic to D…” [27], and it yields
qualitatively identical results with D – type tests, as demonstrated in Table 9, it does not
provide a quantitative measure of H-hypothesis significance. Certain significances tests
related to rs are described briefly in the Appendix.
If the two-factor cases considered here were known to originate (at least approximately)
from a normal population, the ‘standard” randomized block experiment approach would be
admissible for testing the significance of the block effect. A detailed discussion of this
technique, widely documented in the statistical textbook literature, is omitted. Table 10
indicates the possibility of drawing qualitatively identical inferences from nonparametric and
conventional analysis of variance, even if only one of the two is correct, in principle.
104 Thomas Z. Fahidy
*
from ANOVA tables in EXCEL (Windows XP Professional), for randomized block experiments. Blocks are
from top down: sinapic acid; salt types; major producers.
A major fallacy is made when observations obeying a known physical law are subjected
to trend-oriented tests, but without allowing for a specific behaviour predicted by the law in
certain sub-domains of the observation set. This can be seen in Table 11 where a partial set of
classical cathode polarization data has been reconstructed from a current versus total
polarization graph [28]. If all data pairs were equally treated, rank distribution analysis would
lead to an erroneous conclusion, inasmuch as the (admittedly short) limiting-current plateau
for cupric ion discharge, albeit included in the data, would be ignored. Along this plateau, the
independence of current from polarization potential follows directly from the theory of
natural convection at a flat plate, with ample empirical support from electrochemical mass
transport experiments.
Table 11.Analysis of association based on a partial set of cathode polarization data from
experiments with an aqueous 0.732 mol dm-3 CuSO4; 1.484 mol dm-3 H2SO4 solution,
reconstructed from Fig.6 [27]
More generally, testing for a trend is a self-defeating exercise, if the observation sets are
not connected by causality. This cardinal rule, applying equally to parametric and
nonparametric tests, has been amply emphasized in the literature.
While D-statistic based analyses are arguably straightforward, other alternatives, based
e.g. on contingency tables [29,30] are also useful means of analysis, but beyond the scope of
the current presentation. Their potential usefulness to electrochemical systems, not repeated
here, has recently been documented [8].
IV. Conclusion
The preceding applications furnish a small, albeit representative sample of a
nonparametric treatment of electrochemical observations when their probabilistic properties
are unknown, or if no specific a-priori probability distribution can be associated with them.
D-statistic based techniques have much to offer to the electrochemical process analyst, but a
full exploration of this useful tool remains a subject of future research.
106 Thomas Z. Fahidy
Acknowledgments
Facilities for this work were provided by the University of Waterloo and the Natural
Sciences and Engineering Research Council of Canada (NSERC).
Appendix
A Brief Summary of Some Significance Tests Related to Spearman’s Rank
Correlation Coefficient
References
[1] Kraft, C. H., van Eden, C; A Nonparametric Introduction to Statistics; Macmillan:
New York, NY,1968.
[2] Mosteller, F., Rourke, R.; Sturdy Statistics; Addison-Wesley:Reading, MA, 1973.
[3] Hollander, M., Wolfe, D.; Nonparametric Statistical Methods; Wiley: New York, NY,
1973.
[4] Lehman, E.; Nonparametrics: Statistical Methods Based on Ranks; Holden Day: San
Francisco, CA, 1975.
[5] Daniel, W.; Applied Nonparametric Statistics; Houghton, Miffin: Boston, MA, 1978.
[6] Conover, W. J.; Practical Nonparametric Statistics; J. Wiley and Sons: New York,
NY, 3rd edn., 1999.
[7] Higgins, J. J.; An Introduction to Modern Nonparametric Statistics; Duxbury,
Thomson Brooks/Cole: Belmont, CA, 2004.
The Application of D-Statistics Based Tests of Randomness, Independence… 107
Chapter 4
Abstract
Block copolymers have attracted much attention because of their novel properties and various
promising potential applications. However, it is still difficult, if not impossible, to prepare
multiblock copolymers with a controllable chain sequence and block length even though a
variety of synthetic methods, such as anionic and controlled free radical living polymerization
have been advanced. In recent years, we have proposed and developed a novel method of
using the self-assembly of A-B-A triblock copolymers in a solvent which is selectively good
for the two A-blocks. Such self-assembly concentrates and exposes the active groups attached
on the two A-block ends so that they can be coupled together to form a long multiblock
copolymer chain with its sequence and block length controlled by the initial triblock
copolymer. In this review, we first illustrate how the SAAP concept was developed and
exemplified in some real copolymer systems. Furthermore, we compare the coupling
efficiency with and without the self-assembly, and demonstrate that SAAP provides an
elegant way to prepare long multiblock copolymers.
110 Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.
Introduction
In the last two decades, block copolymers have attracted much attention in both polymer
chemistry and physics because of their synthetic challenges and rich phase diagrams in bulk
as well as in solution.[1-13] The preparation of well-defined and complex architectures of
block copolymers requires a chain-growth polymerization mechanism that can be conducted
in the absence of undesired transfer and termination steps.[14] Living polymerization,
especially living anionic polymerization (LAP), is the most powerful tool for the preparation
of block copolymers with designed block lengths and architectures. Chemical modification
after polymerization[15-17] is one of three general strategies to overcome the primary
limitation of using LAP in block copolymer synthesis; namely, the restriction to a minority of
all interesting or otherwise useful monomers. Protection of the functional groups in
monomers is the second approach to extend the application of LAP in the preparation of
block copolymers.[18] The third strategy, combined alternative living or controlled
polymerization protocols with LAP, has been successfully demonstrated.[19-25] Recently,
due to the development of controlled living free-radical polymerization, using one block as a
macroinitiator for another monomer has also been developed.[22, 24, 25]
For preparation of multicomponent block copolymers, it is difficult, if not impossible, to
sequentially polymerize more than five blocks by a single mechanism.[26] A combination of
different mechanisms thereby has provided an alternative route and become more
attractive.[14] In principle, one could sequentially add different types of monomers into a
living anionic polymerization system to prepare long and multiblock copolymer chains.
However, in reality, each addition of new monomer will inevitably terminate some of the
living chain ends because of introducing the impurities,[27] and result in a broad distribution
of the final block copolymer chains. Practically, it is not feasible to fractionate these
multiblock chains because their physical properties, such as solubility and viscosity, are
similar. On the other hand, each sequentially added monomer must be reactive to initiate the
following monomers so that the chain can propagate again.[28] However, in most cases, a
living A block can initiate comonomer B, but a living B block may not initiate comonomer A.
This is the crucial reason why the sequential addition method can only be used to make
copolymers with a few blocks, typically diblock or triblock copolymer chains.[26, 29-31]
Alternatively, one could prepare multiblock copolymers by directly coupling different
blocks with two reactive groups at the end, like the step-growth polymerization. Preparation
of polyurethane is a typical example. However, it should be addressed that such a coupling
reaction is extremely ineffective, especially when long polymer blocks (Mw > 104 g/mol) are
used because most of the reactive ends are wrapped and hidden inside the long coiled blocks
in solution. Moreover, the concentration of the active end is low because of the long chains
used in the coupling reaction. Therefore, it still remains a challenge in polymer chemistry to
synthesize long multi-block heteropolymer chains with an ordered sequence and controllable
block lengths.
Despite the difficulties, the significance of producing such long multiblock
heteropolymers could not be overlooked because they will not only provide unique systems
for the study of polymer physics, but also could lead to a new type of polymeric materials.
The key to a successful preparation of long multiblock copolymer chains is how to effectively
couple the active ends together. It has been well known in polymer physics that the self-
Self-Assembly Assisted Polypolymerization (SAAP) 111
assembly of block copolymer chains in a selective solvent can form different core–shell
micelle-like structures.[27, 32-34] Using such self-assembly, one should be able to force the
reactive ends to stick out and concentrate on the periphery of the micelle-like structure if the
reactive functional groups are attached to the ends of the soluble blocks of copolymer chains.
A combination of the self-assembly and the condensation polymerization (coupling) has
enabled us to propose a new method, the self-assembly assisted polypolymerization (SAAP),
for the preparation of long multiblock copolymers with an ordered chain sequence and
controllable block lengths. Figure 1 schematically illustrates the principle of SAAP.[35]
In the past several years, we have used the SAAP method to prepare different long
multiblock copolymers. Generalities about the preparation and characterization of different
end-functionalized triblock copolymers are first outlined. Then, the micellization of triblock
copolymer as well as coupling efficiency with and without self-assembly method are
discussed.
The peak in f(Rh) located at 3-4 nm represents individual triblock copolymer chains. At 29
°C, an additional peak appears indicating the self-assembly of the triblock copolymer chains.
Pentanediol (HO(CH2)5OH) was added as the linking agent to couple each two functional
ends of the triblock copolymer chains in the presence of pyridine. The resultant multiblock
heteropolymer chains have a structure like (PMMA-b-PS-b-PMMA-c-)n, where “c” denotes
the linking agent, pentanediol. The structure can also be written as (PMMA-b-PS)n, in which
the PMMA block is twice longer than that in the initial triblock PMMA-b-PS-b-PMMA
copolymer chain because each two PMMA blocks are connected together in the resultant
multiblock copolymer.
o
T = 45 C
3.0
f(Rh)
o
T = 29 C
2.0
1.0
0.0 0 1 2
10 10 10
Rh / nm
i 2
Figure 2. Typical hydrodynamic radius distributions (f(Rh)) of individual triblock PMMA-b-PS-b-
PMMA copolymer chain end-capped with oxalyl chloride in a solvent mixture of methyl acetate and
acetonitrile (10/1, v/v) at 45 °C and the aggregates formed via the self-assembly of the triblock
copolymer chains at 29 °C, where the triblock copolymer concentration is 1 x 10-4 g / mL.[35]
1.2
0.8
W(M)
0.4
0.0 3 4 5 6 7
10 10 10 10 10
M / (g/mol)
Figure 3. Weight distributions of molar mass of triblock PMMA-b-PS-b-PMMA copolymer (dashed
line) and long multi-block (PS-b-PMMA)n heteropolymer (solid line) measured by size exclusion
chromatography.[35]
peaks is ~1.25. The weight-average molar mass (Mw) of the resultant multiblock copolymer is
4.72 x 105 g/mol, indicating that on average, more than 20 triblock copolymer chains are
coupled together. According to a normal definition of block copolymers, this is a 40-block
copolymer. The results have been confirmed by static laser light scattering. For comparison, a
control experiment was performed at 45 °C at which no self-assembly of the triblock PMMA-
b-PS-b-PMMA chains was detected by LLS. We are not able to detect any long multi-block
heteropolymer chains by both SEC and LLS after a week, indicating that the self-assembly
plays a crucial role for the preparation of long multiblock copolymers.
1 hr UV in n-heptane.
1 hr UV in THF.
SI42
without UV.
V (a.u.)
1hr UV in n-heptane.
without UV.
SI44
14 16 18 20 22
Retention Volume / mL
Figure 4. SEC profiles of SI42 and SI44 before and after 1 h UV irradiation in n-heptane as well as of
SI42 after 1 h UV irradiation in THF without the self-assembly.[46]
Photodimerization of coumarin and its derivatives have been widely investigated and
used.[37-45] It is known that under a UV irradiation with light wavelength longer than 300
nm, coumarin and its derivatives can undergo [2+2] dimerization to form four types of
dimmers depending on the reaction conditions. Importantly, such formed photodimers can be
114 Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.
reversed when irradiated by a shorter UV light (λ < 254 nm). One might therefore expect that
by terminating of each anionic end with such a photoreactive molecule, like 7-
chlorodimethylsilanoxy-4-methylcoumarin, we can couple two triblock copolymer chains via
the [2 + 2] photodimerization through the coumarin groups.[46] Photo-coupling has exhibited
special advantages over the conventional coupling reaction. For example, the curing time is
fast and the rate of coupling is easily controlled. Additionally, it can avoid the problem of
linking agent addition. It is helpful to note that both the addition of insufficient or excessive
amount of the linking agent reduce the coupling efficiency. In order to test our hypothesis
about the photo-induced coupling reaction, we started with a model precursor diblock
polystyrene-b-polyisoprene copolymer which can self-assemble in n-heptane, a poor solvent
for the PS block. The self-assembly assisted photodimerization of such diblock chains
successfully results in novel polystyrene-b-polyisoprene-c-polyisoprene-b-polystyrene (PS-b-
PI-c-PI-b-PS) “triblock” copolymers (B-A-A-B).
Figure 4 shows typical SEC profiles of two PS-b-PI diblock copolymers before and after
the photodimerization reaction.[46] The appearance of a peak after the UV irradiation at a
short retention time with a doubled molar mass indicates the formation of PS-b-PI-c-PI-b-PS
copolymer chains. In contrast, the irradiation of SI 42 in THF without the self-assembly only
leads to a slight increase of the molar mass. It confirms that the self-assembly helps to
concentrate and expose the reactive ends of precursor diblock copolymers on the periphery of
the core-shell micelles, which greatly increases the coupling efficiency of the
photodimmerization reaction between two coumarin end groups.
is no dimer and multi-mer in the final product, indicating that difunctional polystyryllithium
predominantly undergoes the nucleophilic substitution reaction with 1-bromopentane, where
the side reactions are not significant in THF. Many reports have shown that solvents also play
a pronounceable effect on the reactions. For reactions carried out in a mixture of a
hydrocarbon solvent and polar aprotic solvent, such as THF, there has nearly no side
reactions.[47] In contrast, the reaction in a hydrocarbon solvent can often lead to dimer-
terminated living polymers.[51] The Wurtz-type coupling is likely the reaction pathway,
leading to the formation of dimeric products in the reaction of difunctional polystyryllithium
with dihaloalkanes, but it can be suppressed by the addition of an excess amount of
dihaloalkanes. Our result shows that anionic polymerization provides a feasible route to
prepare well-defined polymer chains with halogen end groups.
Difunctional polystyryllithium
terminated by 1-bromopentane.
4
Mn = 4.25 x 10 g/mol
Mw/Mn = 1.02
V (a.u.)
Difunctional polystyryllithium
terminated by methanol as reference.
4
Mn = 4.22 x 10 g/mol
Mw/Mn = 1.02
10 15 20 25 30
Retention Volume / mL
Fi 5
Figure 5. SEC profiles of difunctional polystyryllithium terminated by 1-bromopentane and methanol,
respectively.
Narrowly distributed PI-b-PS-b-PI triblock copolymer chains with both of their ends
capped with bromobutyl groups were prepared by sequential addition of living anionic
polymerization and terminated by excess of 1,4-dibromobutane (PS block: Mw = 3.5 × 103
g/mol; PI blocks: Mw = 3.1 × 103 g/mol; Mw/Mn = 1.12; The degree of end-functionalization
was 92% characterized by 1HNMR). Figure 6 shows the SEC profile of such prepared
triblock copolymer chains. The small but a detectable amount (~5%) of PI-b-PS-b-PI dimers,
PI-b-PS-b-PI-c-PI-b-PS-b-PI, is presumably formed via the Wurtz-type coupling reaction.
116 Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.
Before the coupling reaction, the self-assembly of PI-b-PS-b-PI triblock copolymer chains in
n-hexane was investigated by LLS. Figure 7 shows typical hydrodynamic radius distributions
(f(Rh)) of individual PI-b-PS-b-PI triblock chains in THF, a good solvent for both the PI and
PS blocks, and the core-shell micelles formed via the self-assembly of the triblock copolymer
chains in n-hexane, a solvent selectively good for the PI block. The shifting of the peak from
3.6 nm to 22.8 nm clearly reveals the self-assembly of individual triblock copolymer chains
and the formation of narrowly distributed micelles. The coexistence of two well-separated
peaks indicates that there is an equilibrium between unimers and micelles.
3
5.9 x 10 g/mol
Unimer.
Coupled in n-hexane.
Coupled in THF.
V (a.u.)
4
1.3 x 10 g/mol
4
5.7 x 10 g/mol
16 20 24 28
Retention Volume / mL
Figure 6. SEC profiles of PI-b-PS-b-PI triblock copolymer chains end-capped with butyl bromide group
(SI44) before and after the coupling reaction in n-hexane, with the self-assembly as well as in THF
without the self-assembly.
It should be emphasized once more that the self-assembly concentrates and exposes the
functional ends of precursor A-B-A triblock copolymer chains on the periphery of each
resultant core-shell micelle. Figure 6 shows that in the formation of −(−PI-b-PS-b-PI−)n−
multiblock copolymer chains, the efficiency of the coupling reaction of 1,4-dilithio-1,1,4,4-
tetraphenylbutane (DD2-) and PI-b-PS-b-PI triblock copolymer chain terminated with the
bromobutyl group is significantly increased by the self-assembly, after we compare the SEC
profiles of the PI-b-PS-b-PI triblock copolymer chains before and after the coupling reaction.
With the self-assembly in n-hexane, the coupling results in a mixture of multi-block
copolymer chains, dimers and the precursor triblock chains. The peak value (5.7 × 104 g/mol)
suggested that on average, 10 triblock chains were linked together to form 21-block
copolymer chains with a defined sequence and controlled block lengths. This is reasonable
because the degree of the end-functionalization is ~90%, which means that on average we can
only link ~10 chains together. In order to make longer multiblock copolymer chains, we will
have to increase the efficiency of the end capping.
In contrast, when the coupling reaction was performed in THF without the self-assembly,
the SEC profile contains only two peaks, namely, a “dimer” peak with Mw ≈ 1.2 × 104 and
precursor triblock copolymer peak. It shows that the coupling reaction stops after two triblock
copolymer chains are linked together. This is because for longer copolymer chains (Mw > 104
g/mol), the chain ends are likely wrapped and hidden inside the chains coiled in a good
Self-Assembly Assisted Polypolymerization (SAAP) 117
solvent. Therefore, the self-assembly is the key step in the preparation of long multiblock
chains in SAAP. In practice, we will have to find a better selective solvent to shift the unimer
↔ micelle equilibrium towards the micelle formation and a better method to functionalize the
ends of the precursor triblock copolymer chains.
12
in n-hexane.
in THF.
8
f(Rh)
0 0 1 2
10 10 10
Rh / nm
Figure 7. Typical hydrodynamic radius distributions (f(Rh)) of individual PI-b-PS-b-PI triblock
copolymer chains end-capped with butyl bromide group (SI44) in THF and their self-assembled core-
shell micelle in n-hexane, where C = 1.0 × 10-2 g/mL and T = 25.0 °C.
V (a.u.)
18 20 22 24 26 28
Retention Volume / mL
Figure 8. SEC profiles of the triblock copolymer precursor ended with carboxyl acid group coupled by
SAAP with 1,6-hexamethylenediamine (HMDA) catalyzed by 1,3-dicyclohexylcarbodiimide (DCC) as
well as coupled directly in THF as a reference.
4
Mw1 = 5.96 x 10 g/mol
6
Mw2 = 1.76 x 10 g/mol Peak 1
Peak 2
V (a.u.)
18 20 22 24
Retention Volume / mL
Figure 9. SEC profiles of the triblock copolymer precursor ended with acyl chloride group coupled by
SAAP with 1,6-hexamethylenediamine (HMDA) but without 1,3-dicyclohexylcarbodiimide (DCC).
In the above coupling reaction, an excess amount of DCC was added to increase the
reactivity of the carboxylic acid group. The carboxylic acid groups have to be activated by the
DCC first then form the amide linkage with the HMDA. In order to increase the coupling
efficiency, we convert the carboxylic acid group to acyl chloride to increase its reactivity. In
this way, the α,ω-diacyl chloride terminated PI-PS-PI triblock copolymer chains can be
coupled by the SAAP method with HMDA without DCC. The SEC result of the coupling
product of the α,ω-diacyl chloride terminated PI-b-PS-b-PI triblock copolymer chains, as
shown in Figure 9, clearly reveals that ~30 triblock copolymer chains are coupled together to
form long multiblock copolymer chains, PI-b-(-PS-b-PI-c-PI-b-)29-PS-b-PI, where “c”
represents the linkage unit. If we treat each PI-c-PI as one PI block, the multiblock copolymer
chains can be regarded as a 60-block copolymer. To the best of our knowledge, the maximum
reported number of blocks of multiblock copolymers was 11 up to now.[29, 30] Our results
Self-Assembly Assisted Polypolymerization (SAAP) 119
clearly demonstrate the advantage of SAAP in the preparation of long multiblock copolymer
with a controllable consequence and different block lengths.
Conclusion
We have developed a novel self-assembly assisted polypolymerization (SAAP) method
for the preparation of long multiblock copolymers with a controllable sequence and different
block lengths. The self-assembly of precursor triblock A-B-A copolymer chains in a solvent
selectively good for the A block concentrates and exposes the functional groups at the end of
the A block on the periphery of each self-assembled core-shell micelle. The SAAP method
enables us to synthesize long multiblock copolymer chains with its sequence and block length
well controlled by the initial living polymerization. Using this novel method, we have
successfully prepared 60-block copolymer chains, (PS-b-PI)30. It opens a door for further
studies of effects of comonomer composition and distribution on the chain backbone on its
solution and bulk properties. The success of this novel approach clearly shows how synthetic
chemistry is combined with polymer physics.[74] The problem-driven synthesis will be
critically important in the study of polymer physics.
Acknowledgment
The financial support of the Research Grants Council of the Hong Kong Special
Administration (CUHK4267/00P, CUHK 4209/99P, 2160122, 2060255), NNSFC 29974027,
the special funds for Major State Basic Research Projects (G1999064800), and the CAS Bai
Ren Project is gratefully acknowledged. Dr. F. M. Zhu wishes to thank Lingnan Foundation
of Zhongshan University and the Postdoctoral Fellowship of the Chinese University of Hong
Kong.
References
[1] Discher, B. M.; Won, Y. Y.; Ege, D. S.; Lee, J. C. M.; Bates, F. S.; Discher, D. E.;
Hammer, D. A. Science 1999, 284, 1143.
[2] Discher, D. E.; Eisenberg, A. Science 2002, 297, 967.
[3] Dubertret, B.; Skourides, P.; Norris, D. J.; Noireaux, V.; Brivanlou, A. H.; Libchaber, A.
Science 2002, 298, 1759.
[4] Jain, S.; Bates, F. S. Science 2003, 300, 460.
[5] Jenekhe, S. A.; Chen, X. L. Science 1999, 283, 372.
[6] Jeong, B.; Bae, Y. H.; Lee, D. S.; Kim, S. W. Nature 1997, 388, 860.
[7] Li, Z. B.; Kesselman, E.; Talmon, Y.; Hillmyer, M. A.; Lodge, T. P. Science 2004, 306,
98.
[8] Russell, T. P.; Karis, T. E.; Gallot, Y.; Mayes, A. M. Nature 1994, 368, 729.
[9] Savic, R.; Luo, L. B.; Eisenberg, A.; Maysinger, D. Science 2003, 300, 615.
[10] Seul, M.; Andelman, D. Science 1995, 267, 476.
120 Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.
[11] Stoykovich, M. P.; Muller, M.; Kim, S. O.; Solak, H. H.; Edwards, E. W.; de Pablo, J.
J.; Nealey, P. F. Science 2005, 308, 1442.
[12] Zhang, L. F.; Eisenberg, A. Science 1995, 268, 1728.
[13] Zhang, L. F.; Yu, K.; Eisenberg, A. Science 1996, 272, 1777.
[14] Lodge, T. P. Macromol. Chem. Phys. 2003, 204, 265.
[15] Ren, Y.; Lodge, T. P.; Hillmyer, M. A. Macromolecules 2000, 33, 866.
[16] Ren, Y.; Lodge, T. P.; Hillmyer, M. A. Macromolecules 2001, 34, 4780.
[17] Hillmyer, M. A.; Schmuhl, N. W.; Lodge, T. P. Macromol. Symp. 2004, 215, 51.
[18] Mori, H.; Hirao, A.; Nakahama, S. Macromolecules 1994, 27, 35.
[19] Lu, Z. J.; Huang, X. Y.; Huang, J. L.; Pan, G. Q. Macromol. Rapid Commun. 1998,
19, 527.
[20] Ramakrishnan, A.; Dhamodharan, R. J. Macromol. Sci.-Pure Appl. Chem. 2000,
37, 621.
[21] Lu, G. Q.; Jia, Z. F.; Yi, W.; Huang, J. L. J. Polym. Sci. Polym. Chem. 2002, 40, 4404.
[22] Bernaerts, K. V.; Du Prez, F. E. Prog. Polym. Sci. 2006, 31, 671.
[23] Miura, Y.; Yamaoka, K.; Mannan, M. A. Polymer 2006, 47, 510.
[24] Yagci, Y.; Tasdelen, M. A. Prog. Polym. Sci. 2006, 31, 1133.
[25] Hadjichristidis, N.; Iatrou, H.; Pitsikalis, M.; Mays, J. Prog. Polym. Sci. 2006, 31, 1068.
[26] Ekizoglou, N.; Hadjichristidis, N. J. Polym. Sci. Polym. Chem. 2002, 40, 2166.
[27] Hadjichristidis, N.; Pispas, S.; Floudas, G. A. Block Copolymers: Synthetic Strategies,
Physical Properties, and Applications John Wiley & Sons, Inc.: Hoboken, N.J., 2003.
[28] Hsieh, H. L.; Quirk, R. P. Anionic Polymerization: Principles and Practical
Applications. Marcel Dekker: New York, 1996.
[29] Nagata, Y.; Masuda, J.; Noro, A.; Cho, D. Y.; Takano, A.; Matsushita, Y.
Macromolecules 2005, 38, 10220.
[30] Wu, L. F.; Cochran, E. W.; Lodge, T. P.; Bates, F. S. Macromolecules 2004, 37, 3360.
[31] Wu, L. F.; Lodge, T. P.; Bates, F. S. Macromolecules 2006, 39, 294.
[32] Hamley, I. W. The Physics of Block Copolymers. Oxford University Press: New York,
1998.
[33] Fredrickson, G. H.; Bates, F. S. Annu. Rev. Mater. Sci. 1996, 26, 501.
[34] Riess, G. Prog. Polym. Sci. 2003, 28, 1107.
[35] Wu, C.; Xie, Z. W.; Zhang, G. Z.; Zi, G. F.; Tu, Y. F.; Yang, Y. L.; Cai, P.; Nie, T.
Chem. Commun. 2002, 2898.
[36] Wu, C.; Xie, Z. W.; Zhang, G. Z.; Zi, G. F. Chin. J. Polym. Sci. 2001, 19, 451.
[37] Kuznetsova, N. A.; Kaliya, O. L. Russ. Chem. Rev. 1992, 61, 1243.
[38] Morrison, H.; Curtis, H.; McDowell, T. J. Am. Chem. Soc. 1966, 88, 5415.
[39] Hoffman, R.; Wells, P.; Morrison, H. J. Org. Chem. 1971, 36, 102.
[40] Muthuramu, K.; Ramamurthy, V. J. Org. Chem. 1982, 47, 3976.
[41] Chen, Y.; Geh, J. L. Polymer 1996, 37, 4473.
[42] Chen, Y.; Geh, J. L. Polymer 1996, 37, 4481.
[43] Ngai, T.; Wu, C. Macromolecules 2003, 36, 848.
[44] Jiang, J. Q.; Qi, B.; Lepage, M.; Zhao, Y. Macromolecules 2007, 40, 790.
[45] Li, W. J.; Lynch, V.; Thompson, H.; Fox, M. A. J. Am. Chem. Soc. 1997, 119, 7211.
[46] Zhu, F. M.; Ngai, T.; Xie, Z. W.; Wu, C. Macromolecules 2003, 36, 7405.
[47] Asami, R.; Takaki, M.; Hanahata, H. Macromolecules 1983, 16, 628.
[48] Hirao, A.; Hayashi, M.; Nakahama, S. Macromolecules 1996, 29, 3353.
Self-Assembly Assisted Polypolymerization (SAAP) 121
[49] Hirao, A.; Tohoyama, M.; Nakahama, S. Macromolecules 1997, 30, 3484.
[50] Hirao, A.; Hayashi, M. Acta Polym. 1999, 50, 219.
[51] Quirk, R. P.; Han, K.; Lee, Y. Polym. Int. 1999, 48, 99.
[52] Bailey, W. F.; Patricia, J. J. J. Organomet. Chem. 1988, 352, 1.
[53] Ashby, E. C.; Sun, X. J.; Duff, J. L. J. Org. Chem. 1994, 59, 1270.
[54] Tillman, E. S.; Hogen-Esch, T. E. J. Polym. Sci. Polym. Chem. 2002, 40, 1081.
[55] Zhu, F. M.; Hong, L.; Li, X.; Wu, C. Manuscript in Preparation.
[56] Szwarc, M. Nature 1956, 178, 1168.
[57] Morton, M. Anionic Polymerization: Principles and Practice. Academic Press: New
YorK, 1983.
[58] Jagur-Grodzinski, J. J. Polym. Sci. Polym. Chem. 2002, 40, 2116.
[59] Quirk, R. P.; Yin, J.; Fetters, L. J. Macromolecules 1989, 22, 85.
[60] Quirk, R. P.; Yin, J.; Fetters, L. J.; Kastrup, R. V. Macromolecules 1992, 25, 2262.
[61] Pan, J.; Chen, M. F.; Warner, W.; He, M. Q.; Dalton, L.; Hogen-Esch, T. E.
Macromolecules 2000, 33, 4673.
[62] Choi, S. B.; Han, C. D. Macromolecules 2003, 36, 6220.
[63] Ji, H. N.; Nonidez, W. K.; Advincula, R. C.; Smith, G. D.; Kilbey, S. M.; Dadmun, M.
D.; Mays, J. W. Macromolecules 2005, 38, 9950.
[64] Sheehan, J. C.; Hess, G. P. J. Am. Chem. Soc. 1955, 77, 1067.
[65] Tang, P. Org. Synth. 2005, 81, 262.
[66] Kubo, M.; Hayashi, T.; Kobayashi, H.; Tsuboi, K.; Itoh, T. Macromolecules 1997, 30,
2805.
[67] Pantazis, D.; Schulz, D. N.; Hadjichristidis, N. J. Polym. Sci. Polym. Chem. 2002, 40,
1476.
[68] Huang, H. Y.; Kowalewski, T.; Remsen, E. E.; Gertzmann, R.; Wooley, K. L. J. Am.
Chem. Soc. 1997, 119, 11653.
[69] Huang, H. Y.; Remsen, E. E.; Wooley, K. L. Chem. Commun. 1998, 1415.
[70] Zhang, Q.; Remsen, E. E.; Wooley, K. L. J. Am. Chem. Soc. 2000, 122, 3642.
[71] Ma, Q. G.; Wooley, K. L. J. Polym. Sci. Polym. Chem. 2000, 38, 4805.
[72] Liu, G. J.; Yan, X. H.; Li, Z.; Zhou, J. Y.; Duncan, S. J. Am. Chem. Soc. 2003, 125,
14039.
[73] Hong, L.; Zhu, F. M.; Li, J.; Ngai, T.; Wu, C. Manuscript in Preparation.
[74] Wu, C. Chem. Eng. News 2001, 79, 197.
INDEX
Amsterdam, 41
A analysis of variance, 94, 95, 103
analysts, 94
AA, 37
analytical techniques, 6
absorption, 34
anionic surfactant, 23
access, 32
anions, 87
acetate, 111, 112
anomalous, 47, 55
acetone, 26, 54
ANOVA, 94, 104
acetonitrile, 111, 112
Antarctic, 37, 63
acid, 3, 5, 95, 98, 104, 117, 118
anthropogenic, 62
acidification, 34
application, 50, 57, 81, 93, 100, 110
activation, 42, 72
aprotic, 115
activation energy, 42, 72
aqueous solution, 10, 39, 40, 71
adamantane, 27
Arctic, 10, 63
additives, 36, 41, 48, 117
argon, 51
adsorption, 3, 34, 46, 64, 65, 66, 70, 87, 88, 89
Arrhenius equation, 72
adsorption isotherms, 87, 88
artificial, 81
aerosol, 87
ash, 24
age, 46
ASTM, 64, 89, 90
agent, 9, 72, 112, 114
Atlantic Ocean, 62
aggregates, 112
atmosphere, 23, 34, 64, 71, 72, 76, 77, 78, 79, 80, 81,
AIDS, 6
87, 88
air, 3, 19, 34, 40, 42, 53, 62, 63, 64, 67, 70, 71, 72,
atmospheric pressure, 28
73, 75, 76, 77, 79, 89
atomic force microscopy, 6
Alberta, 43
atoms, 3, 12
alcohol, 27
attention, 2, 15, 18, 109, 110, 114
alkali, 100, 103, 106
attractiveness, 94
alkaline, 70, 83, 102
automata, 17
alloys, 90, 91
automation, 5
alternative, 36, 40, 45, 50, 94, 96, 106, 110
averaging, 17
alternative hypothesis, 94, 96
avoidance, 54
alternatives, 23, 105
alters, 18
aluminum, 6, 65, 73, 74 B
amide, 3, 4, 117, 118
amine, 26, 117 barrier, 27, 36, 53, 65
amines, 34 batteries, 42
amino, 3, 4 behavior, 22, 25, 46, 48, 54, 75, 77, 81
ammonia, 70 Beijing, 90
124 Index
bicarbonate, 87 chemicals, 37
biological, 40 chemisorption, 3
biologically, 38 chemistry,1, 5, 110, 119
biomedical, 5 China, 49, 50, 90, 109
bipolar, 102 Chinese, 42, 49, 109, 119
blocks, 93, 94, 104, 109, 110, 111, 114, 115, 118 chloride, 62, 73, 80, 81, 82, 85, 87, 88, 111, 112, 118
blood, 37 chlorine, 10
boilers, 34 chromatography, 112
Boston, 6, 106 circulation, 23
British, 9, 89 cis, 51
bubble, 19, 23, 49, 53 classical, 95, 104
building blocks, 13 classification, 63, 85
butane, 14, 18, 20 clean energy, 11, 19, 40, 42
butyl ether, 18 cleaning, 75, 81, 82, 87, 89
climate change, 18, 34, 56, 59
climate warming, 41
C climatic factors, 64, 72
clinical, 5
cadmium, 101
closed-loop, 39
calibration, 11, 14
clusters, 50, 71
California, 6
CNTs, 1, 2, 3, 4
calorimetric method, 19
CO2, 10, 20, 34, 35, 36, 38, 40, 41, 42, 46, 52, 54,
calorimetry, 20
56, 58, 117
Canada, 9, 42, 50, 93, 97, 106
coal, 34, 42
candidates, 19
coastal zone, 77, 87
capacity, 10, 11, 21, 22, 33, 37, 55, 57, 87
coil, 117
capillary, 17, 23, 71
Colorado, 15, 20
capital, 19, 33
combustion, 34, 48
carbon, 3, 4, 9, 10, 11, 12, 14, 20, 23, 34, 38, 39, 40,
commercial, 18, 23, 38
43, 45, 46, 47, 48, 52, 53, 54, 58, 59, 73, 84, 114
complications, 10
Carbon, 1, 2, 18, 38, 47, 48, 51, 52, 58, 59, 117
components, 10, 12, 15, 19, 20, 50, 57, 66
carbon atoms, 14
composite, 6
carbon dioxide, 9, 10, 11, 12, 20, 23, 34, 38, 40, 45,
composition, 19, 26, 42, 119
46, 47, 48, 52, 54, 58, 59, 114
compounds, 12, 20, 39, 62, 73, 80, 83
Carbon nanotubes (CNTs), 1, 2
compressibility, 21
carboxyl, 117, 118
compression, 36, 39
carboxylic, 5, 117, 118
computation, 99
Caribbean, 62, 80, 85
computer, 15, 17
carrier, 25, 52
Computer simulation, 41
CAS, 119
concentrates, 109, 116, 119
casting, 3
concentration, 3, 16, 22, 24, 26, 34, 36, 39, 40, 46,
catalysis, 2, 39
72, 81, 87, 88, 101, 102, 110, 112
catalytic, 3
concrete, 75
cathode, 104, 105
condensation, 70, 72, 75, 79, 111
cathode polarization, 104
conductivity, 5, 65, 88
cathodic process, 64, 71
confidence, 94, 95
causality, 105
configuration, 21
cavities, 9, 12, 13
confusion, 99
CD, 42
Congress, 90
cell, 9, 12, 13, 31, 36, 93, 100, 101, 102, 103
construction, 75
CGC, 26
consumption, 15, 16, 23, 27, 29, 45
CH4, 12, 16, 52, 54, 56
contaminants, 62, 64, 65, 70, 71, 72, 73, 76, 77, 83,
chemical, 10, 11, 14, 40, 42, 43, 87
87, 88
chemical engineering, 40
Index 125
contingency, 94, 105 definition, 63, 64, 65, 66, 69, 76, 88, 89, 113
continuing, 10 degree, 17, 32, 94, 102, 103, 115, 116
continuity, 95, 96 Delta, 11, 42, 50
control, 9, 25, 70, 113 density, 17, 21, 22, 24, 26, 32, 33, 36, 37, 94, 99,
controlled, 3, 109, 110, 114, 116, 119 102, 103, 104, 105
convection, 104 Department of the Interior, 58
conversion, 16, 17, 22, 23, 29, 31, 32, 37, 53, 56 deposition, 3, 43, 62, 70, 71, 73, 76, 77, 79, 80, 81,
conversion rate, 31, 56 82, 83, 84, 85, 86, 88, 89, 94
cooling, 16 deposition rate, 62, 71, 73, 77, 79, 80, 81, 82, 83, 84,
copolymer, 109, 110, 111, 112, 114, 115, 116, 117, 85, 86, 88
118, 119 deposits, 11, 44
copolymers, 109, 110, 111, 113, 114, 118, 119 depression, 15, 26, 44
copper, 64, 73, 81, 82, 85 derivatives, 113
core-shell, 111, 114, 116, 117, 119 desalination, 11, 39, 45, 53
correlation, 66, 67, 82, 94, 103, 106 desert, 63
correlation coefficient, 66, 67, 94, 103, 106 detection, 2
corrosion, 61, 62, 63, 64, 65, 66, 68, 69, 70, 71, 72, deviation, 86
73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, dew, 64, 65, 66, 72, 73, 76, 79
86, 87, 88, 89, 90, 93, 94, 95, 98, 99, 102, 103 diaphragm, 100, 106
corrosivity, 64, 71, 72, 81, 91 differential scanning calorimetry, 15
cost-effective, 19 diffraction, 14, 16, 18, 24, 33, 36, 42, 44, 49, 57
costs, 19, 36 diffusion, 30, 33, 64, 65, 71
cotton, 82 diffusivity, 29, 32
coupling 45, 109, 110, 111, 114, 115, 116, 117, 118 dimer, 115, 116, 117
covering, 24, 29 dimeric, 115
CRC, 107 dimerization, 113, 114
critical value, 71, 103, 106 diseases, 6
cross fertilization, 93 dissociation, 18, 25, 42, 43, 44, 45, 46, 48, 49, 52,
cryogenic, 22, 34 53, 54, 55, 58
crystal, 12, 14, 15, 16, 17, 20, 45, 47, 52, 53, 58 distilled water, 87
crystal lattice, 12, 14 distribution, 15, 46, 54, 56, 94, 95, 96, 97, 98, 99,
crystalline, 9, 10, 12 101, 102, 103, 104, 105, 110, 119
crystallites, 18 distribution function, 94
crystallization, 10, 15, 17, 23, 45 dosage, 43, 49
crystallographic, 22 droplets formation, 90
crystals, 17, 23, 25, 27, 34, 37, 43, 47, 57, 58 dry, 3, 62, 63, 65, 70, 81, 82, 83, 85
Cuba, 61, 62, 63, 64, 66, 67, 68, 70, 72, 75, 79, 80, DS-3, 100
82, 86, 89, 90 DSC, 42, 47, 56
curing, 114 duration, 72
curiosity, 9, 10
cycles, 65, 70
cycling, 3 E
cyclohexane, 26, 57
earth, 10, 38, 39, 40, 44, 83, 84
Earth Science, 49
D ecology, 10
economic, 10, 22, 38, 40, 45
Dallas, 51 economy, 11
data set, 93, 94 Eden, 106
dating, 20 effluents, 39
decisions, 103 electrical, 1, 72
decomposition, 10, 11, 18, 19, 22, 23, 25, 34, 36, 39, electricity, 34
40, 41, 42, 45, 46, 48, 57, 58 Electroanalysis, 5, 7
deep-sea, 45 electrocatalyst, 4, 5
126 Index
glycoproteins, 37 hydroxyl, 3, 4
gold, 3, 5, 64, 65 hypothesis, 94, 95, 96, 98, 100, 103, 106, 114
grain, 24, 43
graph, 104, 105
greenhouse, 10, 39, 41, 56 I
groups, 2, 71, 109, 110, 111, 114, 115, 117, 118, 119
ice, 9, 10, 12, 15, 17, 24, 25, 29, 31, 32, 33, 45, 47,
growth, 15, 16, 17, 23, 24, 28, 29, 31, 32, 37, 38, 41,
53, 57, 58
43, 47, 48, 50, 58, 59, 110
identification, 12, 15
growth rate, 28, 30, 32, 50, 58
images, 32
Guangzhou, 109
imaging, 16, 17, 29, 32, 33, 56
Gulf of Mexico, 62, 64, 84
immobilization, 3
impurities, 100, 110
H in situ, 6, 11, 39, 42, 43, 58
inclusion, 9, 12, 31, 37, 64, 75
H2, 10, 12, 34, 36, 104 independence, 93, 94, 100, 101, 103, 104
H-2A, 100 independent variable, 66, 67, 75
Haj, 7 India, 107
halogen, 114 Indian, 34
heat, 15, 18, 19, 22, 23, 24, 25, 44, 45, 50, 57, 67, induction time, 17, 18, 38
69, 75, 77, 79, 80 industrial, 5, 11, 18, 19, 36, 37, 40, 72, 77, 80, 83,
heat transfer, 22, 24, 25, 45, 50 96, 97, 100, 102
heating, 70 industrial application, 19
heptane, 55, 113, 114 industry, 9, 11, 15, 23, 37
herbicide, 2 infancy, 5
heterogeneous, 2, 5, 17, 58 inferences, 103
heteropolymers, 110 infinite, 15
hexane, 14, 48, 116, 117 inhibition, 37, 43, 47, 49, 59, 93, 94, 98
high pressure, 11, 12, 13, 21, 22, 29, 33, 36, 37, 40, inhibitor, 18, 32, 37, 48, 95, 98, 102
42, 47, 49, 51 inhibitors, 15, 17, 20, 37, 38, 40, 43, 49
high resolution, 44 inhomogeneities, 17
high temperature, 62 injection, 23, 37, 56
HIV, 6 insight, 16
Holland, 53 inspection, 102
homogeneous, 17 instability, 10, 54, 55
Hong Kong, 109, 119 integration, 1, 2, 3, 4
host, 9, 12, 21 intensity, 14, 30, 31, 72
House, 11, 44, 45 interaction, 12, 18, 62, 72, 89
humanity, 18 interactions, 3, 12, 21
humidity, 61, 62, 63, 64, 65, 70, 71, 72, 75, 77, 79, interface, 16, 17, 23, 24, 32, 50, 58, 64, 88
84, 89 interference, 29
hybrid, 5, 34, 35 interval, 65, 75
hydration, 20, 42, 53, 58 intrinsic, 17, 25, 42, 45
hydro, 14, 19, 20, 48, 55 Investigations, 50
hydrocarbon, 9, 11, 20, 40, 115 ions, 62, 73, 80, 81, 82, 85, 87, 88
hydrocarbons, 14, 19, 20, 48, 55 iron, 65, 95
hydrodynamic, 112, 116, 117 irradiation, 67, 113, 114
hydrogen, 9, 10, 11, 12, 13, 19, 20, 23, 36, 37, 40, ISO, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73,
42, 44, 45, 48, 49, 52, 53, 55 74, 75, 76, 78, 79, 80, 82, 83, 85, 86, 88, 89, 90,
hydrogen bonds, 12 91
hydrogen gas, 23, 36 isobutane, 44
hydrogen sulfide, 9, 12, 20, 48 isolation, 70
hydrolysis, 2 isopentane, 52
hydrophobic, 12, 18, 32, 44 isotherms, 87
128 Index
isotope, 39, 43
isotropic, 30
M
Italy, 97
macromolecules, 40
magnet, 23
J magnetic, 23, 29, 47
magnetic resonance imaging, 47
January, 53, 91 management, 42
Japan, 23, 33, 40, 52, 56, 97 mantle, 11
Jung, 91 Mars, 11
MAS, 29
mass loss, 81, 82
K mass transfer, 17, 24, 27, 29, 31, 32, 53
measurement, 15, 17, 31, 42, 43, 58
kinetic model, 17 measures, 52
kinetic studies, 15, 17 mechanical, 1, 23
kinetics, 12, 15, 16, 17, 22, 23, 24, 25, 26, 27, 29, 32, median, 94
37, 42, 45, 56, 58 mediators, 5
Korean, 43, 54 melt, 25, 47, 57
krypton, 44 melting, 24, 32, 45, 55, 64
memory, 17, 38, 41, 51, 59
L MES, 33
mesoscopic, 54
Langmuir, 87, 88 metals, 5, 61, 62, 63, 72, 75, 88, 90, 91
laser, 113 meteorological, 72, 76
lattice, 12, 13, 14, 21, 27, 36, 39 methane, 9, 10, 11, 12, 13, 15, 16, 17, 18, 19, 20, 21,
lattices, 10, 12, 21 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 38,
law, 25, 104 39, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52,
leaching, 80, 85 53, 54, 55, 56, 57, 58
lead, 88, 102, 104, 110, 114 methanol, 15, 37, 43, 115
lifetime, 23 methylcyclohexane, 21, 27, 28, 33, 51, 55
light scattering, 113 Mexico, 61, 62, 67, 80, 84, 89
limitation, 29, 110 micelles, 114, 116, 117
limitations, 76, 79, 88 migration, 16
linear, 16, 29, 62, 65, 88 mines, 20
linear function, 65 miniaturization, 6
linear regression, 65 minority, 110
linkage, 117, 118 mirror, 102
liquefied natural gas, 40 mixing, 23, 24, 29, 33, 56
liquid film, 63, 64, 89 model system, 20
liquid interfaces, 45, 47 modeling, 54
liquid phase, 15, 16, 30 models, 11, 15, 25
liquid water, 17, 18, 23, 24, 29, 48, 52 moieties, 2
literature, 20, 27, 36, 87, 103, 105 moisture, 75
lithium, 114 molar volume, 15
LNG, 19, 21, 22, 33, 40 molecular dynamics, 25, 36, 51
London, 43 molecules, 7, 9, 12, 13, 14, 17, 18, 19, 20, 21, 22, 23,
long period, 62 27, 28, 30, 32, 33, 36, 37, 38, 49, 59
long-term, 87 monolayer, 3, 87
low temperatures, 37, 48, 59 monolayers, 7, 65
monomer, 110
monomeric, 3
Monte Carlo method, 17
morphology, 3, 16, 52
Index 129
NA, 18
Na+, 88 P
NaCl, 51, 72, 98
nanometer, 72 palletized, 22
nanotube, 4 paper, 6, 43, 50, 62
nanowires, 5 parameter, 61, 62, 63, 88, 94
National Academy of Sciences, 47 particles, 24, 31, 33, 37, 64, 87
National Research Council, 9 partition, 29
natural, 9, 10, 11, 19, 20, 25, 34, 35, 38, 40, 43, 45, passivation, 70
47, 52, 54, 56, 57, 62, 65, 72, 80, 81, 85, 104 pentane, 14, 48
natural gas, 9, 10, 11, 19, 20, 34, 35, 38, 40, 45, 47, performance, 22, 23, 37
52, 54, 56, 57 permafrost, 10
needles, 47 permeation, 98, 103, 104
nerve, 2 Petroleum, 41, 43, 44, 49, 53
nerve agents, 2 pH, 88
Netherlands, 44, 50 pharmaceutical, 5
New York, 6, 7, 43, 47, 48, 54, 57, 106, 107, 120 phase boundaries, 50
next generation, 41, 45 phase diagram, 20, 26, 41, 49, 110
Ni, 1, 5 phase transformation, 11, 15, 32
Nielsen, 42 Photocatalytic, 56
nitric oxide, 2 physical properties, 58, 110
nitrogen, 9, 13, 45, 46, 48, 52, 54, 58, 59 physico-chemical properties, 65
NMR, 14, 16, 17, 20, 26, 27, 29, 31, 32, 33, 42, 44, physics, 110, 119
51, 53, 54, 56, 58 pipelines, 9, 10, 19, 37, 40, 43, 46, 52
noise, 29 planetary, 22
nonparametric, 93, 94, 95, 102, 103, 105 planets, 11, 19, 49, 55
non-random, 98 plants, 34
normal, 37, 94, 95, 96, 97, 99, 100, 102, 103, 113 play, 11, 33, 65, 81, 87, 115
NPS, 94 PMMA, 111, 112
NS, 104 polar ice, 19
Nuclear magnetic resonance, 42 polarization, 53, 71, 104, 105
nucleation, 15, 16, 17, 28, 29, 32, 37, 38, 47, 48, 51, politics, 53
53, 59 pollutants, 64, 72, 80, 85
nuclei, 15, 43 pollution, 91
polycrystalline, 58
polydispersity, 112
O polyethylene, 18, 37
polyisoprene, 114
observations, 42, 45, 47, 53, 56, 58, 93, 94, 98, 100,
polymer, 37, 110, 114, 115, 117, 119
104, 105, 106
polymerization, 109, 110, 111, 115, 117, 119
offshore, 10, 43, 44
polymerization mechanism, 110
oil, 9, 10, 34, 37, 42, 45, 62
polymers, 115
130 Index
polystyrene, 114 rain, 62, 64, 65, 66, 72, 73, 74, 75, 76, 79, 81, 82, 85,
polyurethane, 110 87, 88, 89
poor, 23, 31, 32, 114 rainfall, 64, 70, 72, 73, 81, 82
population, 94, 103 Raman, 14, 16, 20, 33, 47, 55, 56, 57, 58, 59
pore, 23 random, 94, 98
porosity, 25 range, 44, 64, 105
porous media, 23, 25, 44, 53 reaction mechanism, 114
porphyrins, 1 reaction rate, 71
positive correlation, 106 reaction time, 31, 37
potassium, 88 reactive groups, 110
potential energy, 39 reactivity, 118
powder, 14, 29, 32, 57 reality, 17, 65, 71, 110
power, 23, 29, 34, 38, 42 reasoning, 17
power generation, 34 recovery, 11, 25, 35, 38, 39, 40, 41, 46
power plants, 34, 38, 42 redox, 1, 2, 3, 4, 5
precipitation, 62, 63, 64, 72, 75, 79, 87, 88 reduction, 19, 22, 26, 33, 36, 71, 95
prediction, 15, 17, 27, 41, 45, 90 refrigerant, 39, 57
predictive model, 17 refrigeration, 11
preference, 20, 21 regression, 66, 67, 73, 82
preparation, 110, 111, 113, 114, 117, 119 regulations, 20
pressure, 9, 10, 11, 14, 16, 17, 19, 20, 22, 24, 25, 26, rejection, 95
27, 28, 29, 31, 33, 36, 37, 42, 44, 46, 47, 49, 55, relationship, 39, 40, 44, 62, 65, 69, 72, 80, 88, 100
87, 102 reliability, 15
Pretoria, 1 research, 1, 3, 4, 5, 10, 11, 19, 40, 41, 42, 61, 62, 81,
prevention, 45 82, 105
probability, 93, 94, 96, 97, 98, 100, 104, 105, 106 research and development, 42
probability distribution, 93, 94, 105 Research and Development, 58
probe, 4, 71 researchers, 2, 20
producers, 104 reservation, 55
production, 10, 11, 18, 33, 38, 41, 44, 45, 50, 93, 96, reserves, 9
97, 104 reservoir, 11, 38, 43, 47, 49
productivity, 17 reservoirs, 19, 38, 39, 45
prognosis, 62, 66, 68 resistance, 6, 24, 29, 100
program, 15, 38 retention, 114
promoter, 44 risk, 54
propane, 12, 20, 21, 25, 26, 27, 36, 40, 41, 44, 46, 48 rods, 48
propylene, 26, 54 room temperature, 111
protection, 90 roughness, 3
protein, 18, 38, 50, 59 runaway, 10, 39
protocols, 102, 110 rural, 64, 67, 68, 69, 75, 76, 77, 82, 83, 87, 88
pulp, 39 Russian, 64
P-value, 104 rust, 87
pyrolytic graphite, 3
S
Q
safety, 20, 26, 52
quartz, 64 SAFT, 15, 48
salinity, 62, 71, 80, 82, 83
salt, 65, 71, 81, 88, 90, 98, 104
R salts, 83, 85, 87, 88, 98
sample, 17, 64, 87, 88, 94, 105
radiation, 70
sample variance, 94
radius, 12, 112, 116, 117
Sao Paulo, 90
Index 131
savings, 39 spectroscopy, 4, 14, 16, 17, 20, 24, 29, 41, 53, 56
scanning calorimetry, 16 speculation, 10
scanning electron microscopy, 6, 24 speed, 18, 24, 99
scatter, 103 stability, 10, 13, 18, 21, 22, 26, 32, 49, 52
scattering, 111 stabilize, 13, 21, 26, 34
science, 53, 94 stack gas, 41
scientific, 9, 10, 18 stages, 23, 24, 29, 35, 62
scientists, 11, 39, 40 standard deviation, 86
SDS, 23 statistical analysis, 93
search, 37, 38 statistics, 93, 94, 95
seawater, 11, 39, 53 steady state, 27
SEC, 94, 102, 103, 112, 113, 114, 115, 116, 118 steel, 70, 71, 73, 75, 77, 79, 80, 81, 82, 83, 84, 85,
sediment, 46, 55 86, 87, 88
seeds, 17, 24 stochastic, 17
selecting, 19, 22, 87 storage, 10, 11, 19, 21, 22, 24, 25, 26, 33, 35, 36, 37,
Self, 3, 7, 56, 109 38, 39, 40, 41, 42, 45, 46, 48, 50, 53, 55, 56, 57
self-assembly, 2, 109, 111, 112, 113, 114, 116, 117, strategies, 2, 25, 110
119 strategy use, 4
SEM, 24 streams, 34
sensing, 2, 3 strength, 103
sensitivity, 5, 41 strong interaction, 3
sensors, 1, 5, 64, 72 styrene, 114
separation, 10, 11, 34, 35, 36, 52 substances, 64, 70
series, 53 substitution, 115
shape, 14, 22, 30, 42 substrates, 3
shear, 50 sulfate, 58, 87, 88, 101
Shell, 23, 44 Sulfide, 49, 51, 54
shelter, 79 sulfur, 3, 62, 73, 80, 83
shores, 83 sulphate, 80
sign, 74 summer, 62, 85
signals, 14 Sun, 23, 25, 27, 42, 48, 55, 109, 121
silica, 23, 25, 36, 37, 40, 52 supercritical, 15
silver, 64, 65 suppression, 38
simulation, 11, 16, 17, 25, 33, 36, 38, 51 surface water, 70, 73
simulations, 26, 33, 40, 43 surfactant, 24, 44, 48
sites, 62, 63, 75, 81, 84, 85 surfactants, 23, 55
SO2, 72, 73, 74, 80 surviving, 49
sodium, 23, 58 symbols, 99
sodium dodecyl sulfate (SDS), 23 symmetry, 14, 102
software, 15 synthesis, 22, 23, 27, 34, 49, 110, 111, 119
solar, 22 synthetic, 21, 38, 40, 45, 109, 110, 119
solid phase, 14, 16, 25, 29, 33 systematic, 28, 49
solid-state, 27 systems, 10, 17, 23, 27, 30, 31, 32, 42, 51, 70, 105,
solubility, 23, 28, 29, 32, 33, 110 109, 110
solutions, 43, 58
solvent, 2, 3, 64, 109, 111, 112, 114, 115, 116, 117,
119 T
solvents, 42, 115
technological, 53
sorption, 87
technology, 9, 11, 19, 20, 23, 33, 36, 41, 44, 45, 57,
Spearman rank correlation coefficient, 103
93
species, 2, 3, 4, 5, 9, 13, 42, 101
Teflon, 38
specific surface, 87
spectra, 14, 29
132 Index
temperature, 3, 9, 10, 11, 14, 15, 16, 17, 19, 20, 22,
24, 25, 26, 27, 28, 31, 33, 36, 39, 53, 61, 62, 63,
V
64, 67, 70, 71, 72, 75, 76, 89, 102, 111
values, 16, 27, 39, 61, 63, 64, 65, 67, 70, 76, 79, 81,
territory, 62, 63, 83
82, 83, 84, 85, 87, 88, 96, 97, 98, 99, 101, 102,
test statistic, 103
103
tetrahydrofuran, 26, 36, 46, 47, 57, 59
van der Waals, 10, 12, 14, 15, 20, 21, 58
textbooks, 94
vapor, 15, 28, 39, 55
theoretical, 94, 98
variable, 15, 73, 74, 75, 88, 89, 102
theory, 95, 104
variables, 66, 68, 75, 100, 102
thermal, 17, 21, 24, 25, 31, 39, 45, 46, 51, 56, 70
variance, 94, 96
thermal conduction, 25
variation, 66, 67
thermal energy, 45
velocity, 70, 72
thermal expansion, 21, 51, 56
ventilation, 70
thermodynamic, 11, 12, 20, 21, 22, 52, 54, 57
viscosity, 29, 110
thermodynamics, 14, 15, 20, 26, 27, 32
voltammetric, 3
Thomson, 106, 107
three-dimensional, 13
Ti, 94, 95, 96, 97 W
time, 9, 15, 16, 17, 23, 25, 26, 28, 29, 32, 37, 45, 51,
61, 62, 63, 64, 65, 66, 67, 68, 69, 71, 72, 73, 74, water, 9, 10, 12, 13, 15, 16, 17, 19, 20, 21, 22, 23,
75, 76, 77, 79, 81, 82, 85, 87, 88, 89, 93, 94, 114 24, 25, 26, 27, 28, 29, 31, 32, 35, 37, 38, 39, 41,
time consuming, 94 42, 44, 45, 48, 49, 50, 51, 52, 54, 55, 56, 58, 61,
time periods, 63 64, 65, 66, 70, 71, 72, 73, 76, 87, 88, 89, 102,
Titan, 11, 49, 52, 57 106, 114
trade, 26, 83 weight loss, 69, 79
trade-off, 26 wells, 11
transfer, 4, 5, 50, 110, 114 wet, 32, 62, 63, 65, 70, 85
transformation, 63, 88 wetting, 32, 63, 64
transition, 1, 3, 53, 55, 94 wind, 62, 63, 70, 72
transmission, 10, 19, 23 windows, 70, 75
transport, 2, 4, 5, 11, 17, 19, 22, 23, 24, 40, 57, 71, winter, 38, 62, 85
104
transportation, 10, 11, 37, 45, 52, 63
traps, 5 X
trend, 4, 28, 30, 32, 93, 94, 95, 96, 97, 98, 100, 102,
103, 104, 105, 106 xenon, 44
tsunamis, 10 X-ray, 56, 58
tuberculosis, 6 X-Ray diffraction, 14, 16, 20, 27, 56, 58
tunneling, 5
two-dimensional, 98 Y
UK, 107 Z
uniform, 61, 71
UNIQUAC, 15 zinc, 73
urban, 64, 68, 69, 75, 80, 83 Zinc, 65, 74
users, 19, 22
USSR, 49, 97
UV, 113, 114
UV irradiation, 113, 114