See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8664465

Application of Magnetic Resonance Imaging to the Investigation of the

Diffusivity of 1,1,1,2-Tetrafluorethane in Two Polymers

Article  in  Applied Spectroscopy · April 2004

DOI: 10.1366/000370204322886708 · Source: PubMed

CITATIONS READS

3 113

2 authors:

Manghaiko Mayele Lothar Oellrich

Health Canada Karlsruhe Institute of Technology
8 PUBLICATIONS   65 CITATIONS    57 PUBLICATIONS   579 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Conformation Analysis of Parathyroid Hormone Residues Critical for Bioactivity View project

Book Review View project

All content following this page was uploaded by Manghaiko Mayele on 28 May 2014.

The user has requested enhancement of the downloaded file.

Application of Magnetic Resonance Imaging to the
Investigation of the Diffusivity of 1,1,1,2-Tetra uorethane in
Two Polymers
M . M AYELE and L. R. OELLRICH*
Institut für Technische Thermodynamik und Kältetechnik, Universität Karlsruhe (TH), Kaiserstr. 12, D-76128 Karlsruhe, Germany
In order to evaluate the suitability of a polymer as a sealing ma-
terial for certa in working  uids used in process plants, information
about the  uid diffusivity into the polymer or the polymer perm e-
ability to the  uid is a prereq uisite. The  uid of interest in the
present work is 1,1,1,2-tetra uorethane, CH 2FCF 3 , a partly  uori-
nated hydrocarbon (HFC) commonly known as refrigerant R134a.
HFC s are increasingly used in refrigeration, air conditioning, and
heat pump applications as substitutes for the chloro uorocarbons
(CFCs) or hydrochloro uorocarbons (H CFCs) that are believed to
be resp onsible for ozone depletion in the stratosphere. The polymers
studied were FPM, a per uoroelastomer, and EPDM, an ethylene-
propylene-diene rubber. The study was carried out using magnetic
resonance imaging (M RI). The contact time dependence of diffusion
of the  uid into the polymer, as well as the spatial distributions of
spin–lattice, T 1, and spin–spin, T 2 , relaxation times, were used as
indicators of the in uence of the EPDM matrix on the mobility of
R134a molecules.
Index Headings: Magnetic resonance im aging; Polymers; MRI;
HFC s; Diffusion in polymers.
INT RODUCTIO N
Solvent diffusion in polymers is important in deter-
mining the physical properties of materials that affect
processing, end use, and shelf life. Many aspects of dif-
fusion in polymers have been studied by using indirect
methods such as gravimetric analysis and optical m icros-
copy. It has been shown that magnetic resonance imaging
(MRI) can be applied to investigate the behavior of sol-
Received 25 March 2003; accepted 27 October 2003.
* Author to whom correspondence should be sent. E-mail:
lothar.oellrich@ciw.uni-karlsruhe.de.
F IG . 1. MRI samp le setup.
0003-7028 / 04 / 5803-0338$2.00 / 0
338 Volume 58, Number 3, 2004 q 2004 Society for Applied Spectroscop y
vents imbibed in a polymer.1– 4 MRI has the advantage
that it does not disturb the diffusion process, unlike m ost
other techniques, which require stopping the diffusion
process and destroying the sample. M oreover, any infor-
mation available from classical nuclear magnetic reso-
nance (NMR) spectroscopy, e.g., self-diffusion coefŽ -
cient, D, spin–lattice, T 1 , and spin–spin, T 2 , relaxation
times, may be obtained on a spatial basis by using M RI.
In this work we investigate some characteristic properties
of CH 2FCF 3 (known as R134a) diffusion into two differ-
ent polymers: FPM (per uororubber) and EPDM (ethyl-
ene-propylene-diene rubber) using nuclear spin density
maps and spatial distributions of T 1 and T 2 of the diffus-
ing  uid molecules.
EXPERIMENTAL
M aterials. 1,1,1,2-Tetra uorethane was supplied by
Solvay Fluor & Derivate GmbH. The elastomers FPM
and EPDM were obtained from Freudenberg Gm bH. The
partially  uorinated refrigerant and the rubbers were used
as received without further treatments.
Equipm ent and Sample Preparation. The imaging
system was a Bruker Avance 200 spectrom eter with m i-
croimaging units and ParaVision as the MRI processing
software.
During operation in refrigeration systems, the poly-
mers are in contact with saturated refrigerant as liquid or
vapor. Therefore, the MRI measurements were carried out
at the vapor pressure of R134a at room temperature. For
this purpose, we built a pressure-proof sample setup that
had to meet NMR requirements and Ž t into the imaging
APPLIED SPECTROSCOPY
 F IG . 2. Diffusion front proŽ les and images of R134a in FPM as a function of contact time.
system. The system consisted of a sample tube and a
comm unicating storage buffer (Fig. 1), so that the liquid
HFC could be rem oved after contacting the polymer and
be stored in the buffer. Although this procedure stops the
diffusion process, the desorption of the refrigerant at its
vapor pressure is sufŽ ciently slow to maintain the HFC
concentration distribution in the polym er long enough for
the images to be acquired. Removing the surrounding liq-
uid perm its adjustment of the receiver gain to the level
of the signal intensity of the liquid contained in the poly-
mer rather than to the very strong signal of the surround-
ing liquid, thus substantially improving the signal-to-
noise ratio of the liquid in the polymer. We used 55 mm
3 4 mm 3 2 mm FPM and EPDM rods in this work.
M easurement M ethods. Since FPM does not contain
hydrogen atoms, 1H imaging is suitable for following the
diffusive penetration of the HFC into the polymer. The
only restriction is the signal intensity. This is primarily
APPLIED SPECTROSCOPY 339

F IG . 3. Integral under the diffusion front curves as a function of con-
tact time in the FPM/R134a system .
determined by the concentration of the liquid in the poly-
mer. This depends on the permeability of the polymer to
HFC.
For EPDM, which contains H atoms, 1 H imaging is
problematic. In this case the observed signal intensity can
possibly result from both partners ( uid and polymer). If
the spin–spin relaxation time, T 2, of the H atoms of the
polymer is extremely short, its NM R signal is very broad
and undetectable.5 This situation, which can be assessed
by imaging the polymer alone, can change, however,
when the polymer comes into contact with the refrigerant,
especially if the refrigerant is able to swell the polymer
in such a way that the side chains become mobile. With
such systems one can still get information on the diffu-
sion process using relative measurements. To do so, a
reference (a piece of polymer that does not contact the
 uid) is Ž xed around the sample tube containing the
HFC/polymer system in such a way that it is measured
concurrently and under the same conditions as the sam-
ple. The diffusion process can then be followed using
difference images between the imbibed polymer and the
net polymer (reference) images.
With the system EPDM/R134a there is an additional
measuring possibility. Since R134a contains F atoms and
EPDM does not, the visualization of the penetration of
R134a in EPDM can be accomplished using 19 F instead
of 1 H imaging. For the system FPM/R134a we used the
1
H-M SME (m ulti slice m ulti echo) imaging pulse se-
quence.6 For the system EPDM /R134a, the 1H-GEFI (gra-
dient echo fast imaging) pulse sequence, derived from
FLASH (fast low-angle shot),7 gave us a better signal-to-
noise ratio, as well as a better contrast. In this system we
also used the 19F-MSM E imaging method.
RESULTS AND DISCUSSION
Diffusion of R134 in FPM . Figure 2 contains a series
of images and diffusion front proŽ les of the refrigerant
R134a in FPM at different contact times. All measure-
ments were made at ambient temperature and lasted for
approximately 4 min each for image acquisitions and 30
s for the proŽ les. It can be seen from the Ž gure that the
340 Volume 58, Number 3, 2004
F IG . 4. FPM layer area increase as a function of contact time in the
FPM/R134a system.
diffusion front, and thus the concentration proŽ le, of
R134a in FPM is smooth and obviously advances with
time toward the center of the sample. Such a concentra-
tion gradient of the diffusing liquid is regarded as char-
acteristic of Fick’s diffusion.8 In Case II diffusion, which
is not seen here, a sharp liquid ingress front advances
linearly with time. Fickian diffusion is described by the
classical equation ]C /]t 5 DDC (D is the Laplace oper-
ator), which gives the spatial variation of the concentra-
tion C as a function of time. A plot of the area under the
diffusion front curves (Fig. 3), which represents a mea-
sure of the  uid uptake as a function of the contact time,
shows that after approximately 25 m in the polymer reach-
es saturation. From Fig. 4 it can be seen that the end of
the swelling process, recognizable by the m aximum layer
area increase, correlates well with the saturation point,
whereby the maximum layer area increase of approxi-
mately 25% (related to the area after 2 m in contact time)
is also reached at approximately 25 min contact time. The
layer area was calculated from simple measurements of
length and width of the layer using PV Wave software.
Diffusion of R134a in EPDM . 1 H Imaging. Figure 5
shows the contact time dependence of the diffusion of
R134a in EPDM . Except for the reference image (Fig.
5a), in which EPDM inside and outside the sample tube
is R134 free, Figs. 5b–5k were obtained from the differ-
ence between images of EPDM imbibed with R134a and
the reference image. The neat integral intensity of R134a
in EPDM is plotted in Fig. 6 as a function of the contact
time. It can be seen from Fig. 6 that the diffusion of
R134a in EPDM occurs Ž rst exponentially with the con-
tact time. Up to a contact time of approximately 2 h the
 uid uptake is sm all. Afterwards the R134a concentration
rises rapidly to an equilibrium value. The polymer is sat-
urated at this time.
19 F-Imaging. The fact that only R134a contains  uo-
rine atoms in the system EPDM /R134a makes it possible
to follow the diffusion process using 19F imaging. We
applied this m ethod using a home-made 19 F Birdcage
probe. Figure 7 shows the results at 4, 10, and 26 min
contact time. With this m ethod we investigated the spatial
distribution of spin–lattice, T 1, and spin–spin, T 2, relax-
ation times of R134a in saturated EPDM. The NMR pa-
F IG . 5. Diffusion of R134a in EPDM. (a) Reference (virgin EPDM samp le). (b–k) Difference images (EPDM with R134a contact minus reference).
Contact time (in minutes) (a) 0; (b) 20; (c) 60; (d ) 80; (e) 110; (f) 130; (g) 150; (h) 173; (i) 199; (j) 219; and (k) 240.

ram eters T 1 and T 2 of a  uid diffusing into a polymer
matrix are important in specifying the motion of the  uid
molecules and the degree of interaction with their envi-
ronment. Generally, strong interactions of  uid molecules
with the polymer side chains result in slower molecular
motion and thus in lower T 1s and T 2 s values. The spatial
distributions of T 1 s and T 2 s indicate the spatial depen-
dence of the in uence of the polym er m atrix on the m o-
bility of the  uid molecules. Thus, the structural char-
acteristics (homogeneity, heterogeneity) of the polymer
can be inferred from T 1 and T 2 distributions. For the de-
termination of T 1 a series of spin density images need to
be acquired, whereby the time between successive ac-
quisitions (repetition time, TR) is varied at a constant
echo time (the time required to receive an NM R signal
in an imaging pulse sequence, TE). In the image series

APPLIED SPECTROSCOPY 341


F IG . 6. Integral intensity of R134a in EPDM as a function of contact
time.
the intensity (the magnetization M 0) is then attenuated
according to Eq. 1:
M 5 M 0 [1 2 exp(2TR/T 1 )]
Applying the image sequence analyzer (ISA ) tool from
Bruker, the series of images is Ž t to Eq. 1 on a pixel-by-
pixel basis using a nonlinear-least-square m ethod. The
contrast of the resulting image is based solely on the T 1 s
of the system, providing a visual representation of the
spatial distribution of T 1s. Figure 8 shows the obtained
T 1 distribution image of R134a in EPDM . It can be seen
from the T 1 proŽ les (above and right of the image) that
there are no large irregularities in the T 1 distribution.
Thus, the polymer matrix can be assumed to be structur-
ally homogenous at the limit of the spatial resolution of
the images. For the two-dimensional sequences we used
a square Ž eld of view (FOV) of 15 mm, a slice thickness
of 5 mm, and a matrix size of 128 3 128 pixels corre-
sponding to a voxel size of 0.117 mm 3 0.117 m m 3 5
mm . This m eans that every pixel in the image displays
a grey-scale value that corresponds to the signal intensity
from a sample voxel with these dimensions. Choosing the
matrix size of 128 3 128 and increasing slice thickness
to 5 mm yielded an image with lower spatial resolution,
but higher signal from each voxel. The T 2 distribution is
determined similarly to the T 1 distribution, whereby TE
F IG . 7. 19
F images of the diffusion of R134a in EPDM at (reading from left to right) 4, 10, and 26 min contact time.
342 Volume 58, Number 3, 2004
is varied at constant TR. The image intensities are Ž t to
Eq. 2:
M 5 M 0 exp(2TE/T 2) (2)
As expected, the T 2 distribution image (not shown here)
also proved to be homogeneous. The Ž t of the integral
intensities of the image series to Eq. 1 (T 1 series) and Eq.
2 (T 2 series) yielded m ean relaxation times T 1 and T 2 of
464 and 4.2 m s, respectively, in the system EPDM /R134a
(Figs. 9 and 10). Contrary to solid systems in which T 1
and T 2 are ordinarily different, in neat liquids T 1 and T 2
are usually approximately equal. The observed large dif-
ference between T 1 and T 2 of the imbibed R134a mole-
cules is indicative of a strong in uence of the polymer
side chains on the m obility of the  uid.
CONCLUSION
The results presented in this work demonstrate the use-
fulness of M RI for probing the molecular diffusivity of
refrigerants in potential candidates for elastomeric seal
materials. This technique allows the  uid diffusion front
progression into the polymer to be noninvasively record-
ed, visualized, and analyzed. From the investigation of
the systems FPM /R134a and EPDM/R134a, we found
that the two polymer probes behave differently in contact
(1) with the refrigerant even if the diffusive process follows
Fick’s law. W hile the saturation point in FPM occurs after
25 min contact time, the EPDM saturation was reached
after approximately 150 min. We applied 19 F imaging to
the system EPDM/R134a to investigate the spatial distri-
bution of the NM R parameters T 1 and T 2, which are in-
dicative of the degree of interaction between the  uid
molecules and the polymer matrix. It was found that both
parameters are homogenously distributed in the polymer
and thus concluded that the EPDM m atrix structure is
isotropic on the scale of these experiments. These results
constitute the Ž rst step in a series of MRI studies on the
diffusivity of HFCs in different polymer/HFC combina-
tions aiming at the determination of the suitability of a
given polymer as sealing material for a given HFC. The
work is in progress in our laboratory.
ACK NOW LEDGM ENTS
The authors thank the Deutsche Forschungsgemeinschaft, DFG, for
supporting this work within the DFG Research Group (Forschergruppe)
FOR 338: Application of NMR to the investigation of structure and
transport processes in disperse system s. We also thank M r. Joachim
Tillich for support with the MRI experiments and M r. Karl-Heinz Rep-
ple for preparation of the MRI probes.
 F IG . 8. T 1 distribution in the EPDM/R134a system .

F IG . 10. M ean T 2 in the EPDM /R134a system.

F IG . 9. M ean T 1 in the EPDM /R134a system.

1. L. A. Weisenberger and J. L. Koenig, Appl. Spectrosc. 43, 1117
(1989).
2. R. Sackin, E. Ciampi, J. Godward, J. L. Keddie, and P. J. McDonald,
Macromolecules 34, 890 (2001).
3. I. Hopkinson, R. A. L. Jones, P. J. M cDonald, B. New ling, A. Lecat,
and S. Livings, Polymer 42, 4947 (2001).
4. P. MansŽ eld, R. Bowtell, and S. Blackband, J. Magn. Reson. 99, 507
(1992).
5. W. P. Rothwell, D. R. Holecek, and J. A. Kershaw, J. Polym. Sci.:
Polym. Lett. 22, 241 (1984).
6. L. A. Weisenberger and J. L. Koenig, Macromolecules 23, 2445
(1990).
7. A. Haase, J. Frahm, D. Matthaei, W. Hänicke, and K.-D. Merboldt,
J. Magn. Reson. 67, 258 (1986).
8. M . Valtier, P. Tekely, L. Kiéné, and D. Canet, Macomolecules 28,
4075 (1995).

APPLIED SPECTROSCOPY 343

View publication stats