Synthesis of perfluorinated functionalized, branched ethers

Article abstract of DOI:10.1016/j.jfluchem.2003.12.012

We wish to report synthesis of perfluorinated functionalized, branched ethers from their hydrocarbon analogues by direct fluorination. Yields up to 90%, with high purities, have been obtained at ambient temperature and pressure. This technique will likely develop into a new general method for producing perfluorinated, hyperbranched and dendritic polymers.

Full text of DOI:10.1016/j.jfluchem.2003.12.012

Journal of Fluorine Chemistry 125 (2004) 749–754
Synthesis of perfluorinated functionalized, branched ethers
Kyle W. Fellinga,*, Cameron R. Youngstromb, Richard J. Lagowb
aDepartment of Chemistry and Chemical Engineering, South Dakota School of Mines and Technology,
501 East St. Joseph Street, Rapid City, SD 57701-3995, USA
bDepartment of Chemistry and Biochemistry, College of Natural Sciences, Mail Code A5300,
The University of Texas at Austin, Austin, TX 78712-1167, USA
Received 11 December 2003; received in revised form 15 December 2003; accepted 17 December 2003
Available online 21 January 2004
Abstract
We wish to report synthesis of perfluorinated functionalized, branched ethers from their hydrocarbon analogues by direct fluorination.
Yields up to 90%, with high purities, have been obtained at ambient temperature and pressure. This technique will likely develop into a new
general method for producing perfluorinated, hyperbranched and dendritic polymers.
# 2004 Elsevier B.V. All rights reserved.
Keywords: Direct fluorination; Perfluorinated ethers; Hyperbranched polymers; Dendrimers
1. Introduction
examples in the literature of synthesized fluorodendrimers
[13,14]. Of these, the most common is that which has a
hydrocarbon core and a fluoropolyether tail. These have
been used extensively in supercritical CO2 applications as
surfactants [15]. Other dendrimers of interest were synthe-
sized based on fluorophenylene units [16]. When synthe-
sized, these molecules formed mesogenic liquid–crystalline
phases [17]. To our knowledge, there have been no reports in
the literature to date of any dendritic structures in which
perfluorinated units form the core structure.
Utilizing solution-phase direct fluorination techniques, our
research group has synthesized and perfluorinated branched,
multi-functional compounds which are capable of being
branched into novel dendritic structures. These compounds
are some of the first dendritic monomers reported which are
completely perfluorinated. These compounds might become
useful in supercritical CO2 applications or electronic devices.
Over the years, many different fluoropolymers have been
synthesized. Beginning with poly(chlorotrifluoroethylene)
in 1934 [1], to the accidental discovery of Teflon by Plunkett
[2], many of these polymers have been found to be useful in
industry. These uses include lubricants, insulators, bioma-
terials, coatings, and low surface tension additives [2–5].
The synthesis of these polymers typically consists of
polymerization of fluorine-containing monomers. These
synthesized polymers usually possess many unique thermal,
chemical, and physical properties [3].
Direct fluorination, both solid and solution phase, has been
used to fluorinate many types of polyethers [6–8]. Of note, is
the lack of functionality outside of the ether linkages them-
selves present in the perfluorinated analogues. Functionali-
zation of organofluorine compounds often involves high cost
starting materials and a number of inefficiencies. Methods
used for producing difunctional fluorocarbon ethers are
electrochemical fluorination [9], direct fluorination [10],
hexafluoropropylene oxide reactions with diacyl fluorides
in the presence of alkali metal catalysts [11], and dimeriza-
tion of perfluoroether acid fluorides by photolysis [12].
The area of fluorinated dendrimers is a relatively new field
in fluorine chemistry. There have only been a few reported
2. Results and discussion
Using the Exfluor–Lagow fluorination process, functiona-
lized, branched ethers have been fluorinated with high yields.
Both a general scheme (Schemes 1 and 2) and a summary
(Table 1) of the performed fluorination reactions are given.
A multi-functional, branched compound has been synthe-
sized based on a pentaerythritol core (Scheme 3). This
compound can have the potential use as a dendritic monomer.
* Corresponding author. Tel.: þ1-605-394-2491; fax: þ1-605-394-1232.
E-mail address: kyle.felling@sdsmt.edu (K.W. Felling).
0022-1139/$ – see front matter # 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2003.12.012

750
K.W. Felling et al. / Journal of Fluorine Chemistry 125 (2004) 749–754
MeO2C
CO2Me
H2
C
H2C
CH2
C
H2
O
O
CH2 H2C
C
O
O
C
H2C
H2
H2
C
C
H2C
CH2
H2
MeO2C
CO2Me
1. F2/N2
1. F2/N2
2. MeOH
2. H2O
MeO2C
CO2Me
HOOC
COOH
F2
C
F2
C
F2C
CF2
C
F2C
CF2
C
F2
F2
O
O
CF2 F2C
C
O
O
CF2 F2C
C
O
O
C
F2C
O
O
C
F2C
F2
F2
F2
C
F2
C
F2C
CF2
C
F2C
CF2
F2
MeO2C
CO2Me
HOOC
COOH
Scheme 1.
H3CC(O)O
OC(O)CH3
H2
C
n
H2C
CH2
C
n
H2
O
O
CH2 H2C
C
O
O
C
H2C
H2
H2
C
C
H2C
CH2
H2
n
H3CC(O)O
OC(O)CH3
n
1. F2/N2
1. F2/N2
2. MeOH
2. H2O
MeO2C
CO2Me
HOOC
COOH
F2C
O
O
CF2
F2C
O
O
CF2
F2
C
F2
C
n
n
F2C
CF2
C
F2C
CF2
C
n
n
F2
F2
O
O
CF2 F2C
C
O
O
O
CF2 F2C
C
O
O
C
F2C
O
C
F2C
F2
F2
F2
C
F2
C
C
F2C
CF2
C
F2C
CF2
F2
F2
F2
C
F2
C
n
n
O
O
CF2
O
O
n
CF2
n
MeO2C
CO2Me
HOOC
COOH
Scheme 2.
Pentaerythritol and acrylonitrile in dioxane/water gives a
tetranitrile by a Michael addition reaction (70% yield).
Esterification is achieved by refluxing in methanol under
dry, acidic conditions (82% yield). For the branched poly-
mers, pentaerythritol ethoxylate is acetylated using acetyl
chloride (Scheme 4). Elemental fluorine is then passed
through a solution of these materials.
Table 1
Yields of perfluorinated functionalized, branched ethers
Compound
Relaxation
Yield (%)
temperature (8C)
C(CF2OCF2CF2CO2Me)4
C(CF2OCF2CF2COOH)4
RT
RT
0
60.2 (isolated)
40.5 (isolated)
>90 (by mass)
>90 (by mass)
C[CF2(OCF2CF2)nOCF2CO2Me]4
C[CF2(OCF2CF2)nOCF2COOH]4
The concentration of the fluorine must be initially kept
low, by having a high dilution rate of N2, to slow the
0

K.W. Felling et al. / Journal of Fluorine Chemistry 125 (2004) 749–754
751
CN
NC
HO
HO
OH
OH
O
O
O
O
CN
NC
CN
HCl (g)
MeOH
CO2Me
MeO2C
O
O
O
O
MeO2C
CO2Me
Scheme 3.
HO
OH
H3CC(O)O
OC(O)CH3
H2
C
H2
C
n
n
H2C
CH2
C
n
H2C
CH2
C
n
H2
H2
O
O
CH2 H2C
C
O
O
O
CH2 H2C
O
O
CH3C(O)Cl
C
C
H2C
O
C
H2C
H2
H2
H2
C
H2
C
C
H2C
CH2
C
H2C
CH2
H2
H2
n
n
HO
OH
H3CC(O)O
OC(O)CH3
n
n
Scheme 4.
fluorination process enough to preserve the carbon–oxygen
bonds. As the molecules incorporate more fluorine, more
rigorous conditions are required to drive the fluorination to
completion. This is achieved by increasing the fluorine
concentration through dropping the rate of addition of N2.
These reactions proceed on a very reasonable timescale.
Addition of the substrate occurs over a time period of 7 h
whereupon the reaction is allowed to proceed for 1 h longer.
There is also no need to cool the reactor for the esters as
better yields are obtained when the reactor is left at room
temperature. This leads to less overall materials consump-
tion (liquid nitrogen and elemental fluorine). A slight reduc-
tion in temperature and longer times are needed for the
fluorination of the acetyl materials.
The branched compounds are perfluorinated as they are
exposed to the elemental fluorine in the reactor. This
includes the ester functionality. These compounds are quite
stable if kept in an ambient, inert atmosphere. However, on
exposure to moisture from the air, the fluoroester function-
ality rapidly degrades to give a perfluorinated carboxylic
acid [18]. In order to quench this fluoroester functionality,
methanol can be pumped into the reactor after purging. The
methanol acts as a nucleophile to displace the fluoroester
functionality and form a hydrocarbon methyl ester.
The functional groups can also be transformed into a
variety of other moieties, such as: carboxylic acids, carbox-
ylates, amides, chlorides, hydrides, alcohols, ethyl thioe-
sters, cyanoethyl ethers, hydroxyethyl esters, and acid
chlorides. These conversions have all been shown in detail
elsewhere [3].
These tetrafunctional materials have been supplied to the
research laboratories of Dr. Keith Johnston in the chemical
engineering department of the University of Texas at Austin
for investigation into their utility in supercritical carbon
dioxide. The key to designing surfactants that can stabilize
water-in-CO2 microemulsions is to find a suitable ‘‘CO2-
phillic’’ molecular group that provides a steric barrier to
aggregation. Preliminary results have shown that the di-
functional linear perfluoropolyethers exhibit superior
abilities to form water–CO2 microemulsions [19,20]. The
investigations into the branched materials are ongoing.
Using established organic synthetic techniques; these
compounds can be branched into dendritic structures. These
compounds are a few of the first reported perfluorinated
core, dendritic monomers. Possible applications of these
dendritic materials include uses as micelles and surfactants
in supercritical CO2. Other possible applications include
uses in electronics and possibly liquid–crystalline devices.

752
K.W. Felling et al. / Journal of Fluorine Chemistry 125 (2004) 749–754
3. Experimental
3.1. General
Tetrakis-[(2-cyanoethoxy)methyl]-methane (9.649 g,
0.0277 mol) was dissolved in dry methanol and saturated
with dry hydrogen chloride gas at room temperature.
The solution was refluxed for 3 h under continuous
passing of hydrogen chloride gas. During cooling, argon
was passed through the setup. Hydrolysis was achieved
with ice-cold water. Products were obtained through
extraction with diethyl ether and drying of the organic
layers over sodium sulfate. Distillation of the solvent
gives the methyl ester as a light yellow oil, 10.901 g
(0.0227 mol, 82%).
CIMS (negative mode) m/z (rel. int.) 481 [M þ 1]À (100).
1H NMR (360 MHz, C6D6): d 3.609 (12H, s, H-4), 3.565
(8H, t, J ¼ 6 Hz, H-3), 3.253 (8H, s, H-1), 2.467 (8H, t,
J ¼ 6 Hz, H-2). 13C NMR (90 MHz, C6D6): d 35.0 (s, C-4),
45.4 (s, C-1), 51.8 (s, C-6), 66.6 (s, C-3), 69.6 (s, C-2), 172.5
(s, C-5).
Pentaerythritol, acrylonitrile, pentaerythritol ethoxylate
(3/4 EO/OH), and acetyl chloride were purchased from
Aldrich Chemical Co. (Milwaukee, WI). Elemental fluorine
was purchased from Air Products (Allentown, PA).
Low-resolution mass spectrometry was performed on a
MAT TSQ-70 spectrometer. High-resolution mass spectro-
metry was performed on a VG analytical ZAB2-E mass
1
spectrometer. H and 13C NMR spectra were taken on a
Bruker 360 nuclear magnetic resonance spectrometer. Other
1H, 13C and 19F NMR spectra were taken on a Varian Unity
Plus-300 nuclear magnetic resonance spectrometer. 1H
NMR spectra were recorded at either 300 MHz or
360 MHz using C6D6 as the lock solvent and internal
reference. 13C NMR spectra were recorded at 75 and
90 MHz using C6D6 as the lock solvent and internal refer-
ence. 19F NMR spectra were recorded at 282 MHz using
C6D6 as the lock solvent and CFCl3 as the internal reference.
19F NMR chemical shifts are given in ppm with negative
values denoting resonances to high field of CFCl3. Direct
fluorinations were performed in a similar manner to the
reactions that have been previously described in the litera-
ture [21].
3.1.3. Direct fluorination of tetrakis-([2-(methoxy-
carbonyl)-ethoxy]methyl)methane: perfluorotetrakis-
([2-(methoxycarbonyl)-ethoxy]methyl)methane
The fluorination of tetrakis-[(2-cyanoethoxy)methyl]-
methane was carried out in a solution phase fluorination
reactor that has been described in the literature [21]. The
reactor was filled with 350 ml of 1,1,2-trichlorotrifluor-
oethane (Freon1 113) and 30 g of NaF (0.714 mol) and
kept at room temperature. The reactor was then purged with
N2 (200 cc/min) for 1 h.
3.1.1. Synthesis of tetrakis-[(2-cyanoethoxy)methyl]-
methane
Tetrakis-[(2-cyanoethoxy)methyl]-methane (2.26 g, 4.71
mmol) was dissolved in 150 ml of Freon1 113 inside a
round bottom flask, and then pumped into the purged reactor
at a rate of 25 ml/h. During the addition of tetrakis-
[(2-cyanoethoxy)methyl]-methane to the reactor a N2/F2
mixture was bubbled through the reactor at a rate of
200/50 cc/min, respectively. After the solution containing
tetrakis-[(2-cyanoethoxy)methyl]-methane was completely
pumped into the reactor, the N2/F2 flow rate and the tem-
perature was held constant for an additional hour. The flow
rate was changed to 50/0 cc/min N2/F2, respectively, and the
reactor was allowed to purge for 4 h. After the N2 purge,
50 ml of methanol is pumped into the reactor to quench the
fluoroester functionality and form the methyl ester. The
solution was then filtered to remove any NaF or NaHF2
and the solvent was distilled off.
The following synthesis was adapted from the literature
[22–25]:
Pentaerythritol (12.567 g, 0.0924 mol) was dissolved
in 40 ml dioxane and 2 ml water and mixed with 1 ml
40% aqueous potassium hydroxide solution. The solution
was cooled to freezing in an ice bath. Acrylonitrile
(36.27 g, 0.6843 mol) was added and the mixture stirred
for 48 h at room temperature. The temperature was then
increased to 60 8C for 8 h and allowed to cool to room
temperature. Solvents were evaporated; the remaining oil
was diluted in 100 ml CH2Cl2 and washed with 10%
aqueous sodium chloride solution. The aqueous layer
was re-extracted with CH2Cl2 and the combined organic
layers were dried over sodium sulfate. Solvents were
intensively evaporated by rotary evaporation and product
was obtained as a white crystal, 22.543 g (0.0647 mol,
70%).
Distillation of product (90 8C/100 mmHg) gave a clear,
yellow liquid that was analyzed as C(CF2OCF2CF2-
COOCH3)4 (2.83 mmol, 60.2%). CIMS (negative mode)
m/z (rel. int.) 913 [M þ 1]À (100). Elemental compositions
were studied by high-resolution mass spectroscopy in
chemical ionization mode. Results were consistent
with C(CF2OCF2CF2COOCH3)4 (calculated: 913.288172;
found: 913.287351). 19F NMR (282 MHz, CFCl3): d
À122.663 (s, 8F, –CF2CO–), À86.541 (s, 8F, –OCF2–)
CIMS (negative mode) m/z (rel. int.) 349 [M þ 1]À (100).
1H NMR (360 MHz, C6D6): d 2.591 (8H, t, J ¼ 6 Hz, H-3),
3.477 (8H, s, H-1), 3.651 (8H, t, J ¼ 6 Hz, H-2). 13C NMR
(90 MHz, C6D6) d 19.0 (s, C-4), 45.9 (s, C-1), 65.9 (s, C-3),
68.9 (s, C-2), 118.5 (s, C-5).
3.1.2. Synthesis of tetrakis-([2-(methoxycarbonyl)-
ethoxy]methyl)methane
1
À76.464 (s, 8F, –CF2O–). H NMR (300 MHz, C6D6): d
4.089 (12H, s, H-1) 13C NMR (75 MHz, C6D6): d 53.453
(s, C-6), 158.544 (s, C-5).
The following synthesis was adapted from the literature
[22–25]:

K.W. Felling et al. / Journal of Fluorine Chemistry 125 (2004) 749–754
753
3.1.4. Direct fluorination of tetrakis-([2-
pletely pumped into the reactor, the N2/F2 flow rate changed
to 300/40 cc/min while the temperature was held constant
for an additional hour. The ratio was then changed to 15 and
20 cc/min and the temperature was raised to 5 8C and the
reaction was run for an additional 30 h. Lastly, the tempera-
ture was raised to 30 8C keeping the flow rate constant for
another 14 h. The flow rate was changed to 50/0 cc/min
N2/F2, respectively, and the reactor was allowed to purge for
4 h. The solution was then filtered to remove any NaF or
NaHF2. Five milliliters of water was then added to the
solution and stirred overnight to hydrolyze the perfluori-
nated ester to the perfluorinated acid. The water was sub-
sequently separated and distillation of product gave a clear,
yellow liquid that was analyzed as the perfluoro(tetra(ethy-
lene glycol carboxylic acids)): C[CF2OCF2COOH]4 and
C[CF2OCF2CF2OCF2 COOH]4. The yield was in excess
of 90% (calculated by mass).
(methoxycarbonyl)-ethoxy]methyl)methane: perfluoro-3-
[3-(2-carboxyethyl)-2,2-bis-(2-
carboxyethoxy)methylpropxy]propionic acid
The synthesis of perfluoro-3-[3-(2-carboxyethyl-2,2-bis-
(2-carboxyethoxy)methylpropxy]propionic acid proceeds as
described for tetrakis-[(2-cyanoethoxy)methyl]-methane.
The exception being the addition of the methanol following
fluorination is omitted.
After the distillation of solvent, 10 ml of water is added
to the flask to hydrolyze the compound to the correspond-
ing acid. The product is soluble in the water layer.
C(CF2OCF2CF2COOH)4 (1.85 mmol, 40.5%)was subse-
quently isolated.
CIMS (negative mode) m/z (rel. int.) 857 [M þ 1]À (100).
Elemental compositions were studied by high-resolution
mass spectroscopy in chemical ionization mode. Results
were consistent with C(CF2OCF2CF2COOH)4 (calculated:
856.939779; found: 856.939056). 19F NMR (282 MHz,
CFCl3): d À121.111 (8F, CF2), À85.958 (8F, CF2),
CIMS (negÀative mode) m/z (rel. int.) 656 [M þ 1]À (75),
1120 [M þ 1] (100). 19F NMR (282 MHz, CFCl3): d À74.6
(s, –CF2O–), À76.6 (s, –CF2O–), À77.5 (m, –CF2O–),
À88.5 (m, –CF2O–).
1
À66.549 (8F, CF2). H NMR (300 MHz, C6D6): d 8.159
(4H, s, H-1).
3.1.5. Synthesis of tetra(ethylene glycol acetate)
Acknowledgements
Fifty milliliters of chloroform and 35.6 ml (0.5 mol) of
acetyl chloride were placed in an ice cooled three-necked
flask equipped with an addition funnel and a condenser
under an argon atmosphere. Twenty-five grams of pentaer-
ythritol ethoxylate (3/4 EO/OH) avg. Mn ca 270 (0.11 mol),
dissolved in 40 ml of chloroform, were then slowly added to
the flask with stirring. Upon completion of the addition, the
flask was allowed to warm to room temperature and subse-
quently refluxed overnight. Excess acetyl chloride was
removed under vacuum and the resulting product,
C[CH2(OCH2CH2)nOC(O)CH3]4 was vacuum distilled
resulting in a colorless, viscous liquid.
We thank the Robert A. Welch Foundation and the
National Science Foundation for their financial support of
this research, Grant no. CHE-997288.
References
[1] J. Gadsby, Trans. Plastics Inst. 22 (1954) 5.
[2] R.E., Banks, D.W.A. Sharp, J.C. Tatlow, Fluorine: The First Hundred
Years (1886–1986). Elsevier, Sequoia, 1986, p. 399.
[3] C.R. Youngstrom, Elemental Fluorination: The Synthesis of
Fluorinated Polymers, Organometallics, and Organics Molecules,
Department of Chemistry and Biochemistry, The University of Texas
at Austin, 2002.
CIMS (positive mode) m/z (rel. int.) 480 [M þ 1]þ n ¼ 1
(75), 656 [M þ 1]þ n ¼ 2 (100). 1H NMR (300 MHz,
C6D6): d 2.315 (3H, s, –OC(O)CH3), 3.75 (m, –CH2O–),
3.88 (m, –OCH2–), 4.35 (m, –CH2O–), 4.44 (m, –CH2O).
[4] R.E. Banks, Fluorocarbons and Their Derivatives, second ed.,
Macdonald, London, 1970.
[5] R.E. Banks, M.G. Barlow, Fluorocarbon and Related Chemistry, vol.
2, The Chemical Society, London, 1974, pp. 98–99.
[6] K. Sung, R.J. Lagow, J. Am. Chem. Soc. 117 (1994) 4276.
[7] A.J. Otsuka, R.J. Lagow, J. Fluorine Chem. 4 (1974) 371.
[8] J.L. Margrave, R.J. Lagow, Polym. Lett. 12 (1974) 177.
[9] V.Y. Kazakov, R.A. Dzerhinskaya, V.I. Tsimbalist, E.A. Shishkin, J.
Gen. Chem. USSR 36 (1966) 1799.
3.1.6. Direct fluorination of tetra(ethylene glycol acetate):
perfluoro(tetra(ethylene glycol carboxylic acid))
The fluorination of tetra(ethylene glycol acetate) was
carried out in a solution phase fluorination reactor that has
been described in the literature [21]. The reactor was filled
with 1000 ml of 1,1,2-trichlorotrifluoroethane (Freon1 113)
and 200 g of NaF (4.76 mol) and chilled to 0 8C. The reactor
was then purged with N2 (200 cc/min) for 1 h.
[10] H.C. Wei, Synthesis in Organic Fluorine Chemistry: Fluorocarons,
Polymers, and Their Applications, Department of Chemistry and
Biochemistry, The University of Texas at Austin, 1997.
[11] C.G. Fritz, E.P. Moore, US Patents 3,250,807 and 3,317,484 (1966)
to E.I. du Pont de Nemours and Co.
Tetra(ethylene glycol acetate) (4.00 g, 15 mmol) was
dissolved in 350 ml of Freon1 113 inside a round bottom
flask, and then pumped into the purged reactor at a rate of
25 ml/h. During the addition of tetra(ethylene glycol acet-
ate) to the reactor a N2/F2 mixture was bubbled through the
reactor at a rate of 400/80 cc/min, respectively. After the
solution containing tetra(ethylene glycol acetate) was com-
[12] R.A. Mitsch, P.H. Ogden, A.H. Stoskopf, J. Org. Chem. 35 (1970)
2816.
[13] T.M. Miller, T.X. Neenan, Chem. Mater. 2 (1990) 346–349.
[14] K. Wooley, C. Klug, T. Kowalewski, J. Schaefer, Proc. Am. Chem.
Soc.: Div. Polym. Mater. 73 (1995) 230–231.
[15] A.I. Cooper, J.D. Londono, G. Wignall, J.B. McClain, E.T. Samulski,
´
J.S. Lin, A. Dobrynin, M. Rubinstein, A.L.C. Burke, J.M.J. Frechet,
J.M. DeSimone, Nature 389 (1997) 368–371.

754
K.W. Felling et al. / Journal of Fluorine Chemistry 125 (2004) 749–754
[16] T.M. Miller, T.X. Neenan, R. Zayas, H.E. Bair, J. Am. Chem. Soc.
114 (1992) 1018–1025.
Lagow, S.E. Webber, K.P. Johnston, J. Am. Chem. Soc. 10 (2003)
3181–3189.
[17] A. Pegenau, X.H. Cheng, C. Tschierske, P. Goring, S. Diele, New J.
Chem. 23 (1999) 465–467.
[21] T.R. Bierschenk, T.J. Juhlke, R.J. Lagow, H. Kawa, US Patent
5,093,432 (1992).
[18] K. Murata, H. Kawa, R.J. Lagow, J. Am. Chem. Soc. 120 (1998)
7117.
[22] R. Worl, H. Koster, Tetrahedron 55 (1999) 2941.
[23] H.A. Bruson, US Patent 2,326,721 (1943).
[24] A. Halabi, M.C. Strumia, J. Org. Chem. 65 (2000) 9210.
[25] G.R. Newkome, A. Mishra, C.N. Moorefield, J. Org. Chem. 67
(2002) 3957.
[19] P.A. Psathas, E.A. Sander, W. Ryoo, D. Mitchell, K.W. Felling, R.J.
Lagow, K.T. Lim, K.P. Johnston, J. Disp. Sci. Tech. 23 (2002) 81–92.
[20] C.T. Lee, W. Ryoo, P.G. Smith, J. Arellano, D.R. Mitchell, R.J.