224
S.-z. Zhu et al. / Journal of Fluorine Chemistry 123 (2003) 221–225
From the relative viscosity data, we could conclude that the
4.1. Preparation of 4,40-(hexafluoroisopropylidene)-
bis-(o-xylene) (3)
polyimides 7a0–d0 prepared in DMF have higher molecular
weight than that of 7a–d prepared in xylene. The extent of
polymerization of 6-FDA with 1,2-diaminoethane is better
than the extent of polymerization of 6-FDA with 1,2-diami-
nohexane, while 1,4-phenylenediamine is better than that of
1,3-phenylenediamine. Comparing to the data of [Z], it means
that the raw of polyimide 7c0 which prepared from 6-FDA
with 1,4-phenylenediamine in DMF was highest.
4.1.1. Preparation of 4,40-(hexafluoroisopropylidene)-
bis-(o-xylene) (3) from hexafluoroacetone (2)
An autoclave (100 ml) was charged with o-xylene (10.6 g,
0.1 mol), hexafluoroactone (10 g, 0.06 mol), and HF (22 g,
1 mol). The mixture was heated to 110 8C for 20 h, and
the pressure was about 20 kg/cm2. After cooling, the reac-
tion mixture was poured into ice-water, and the oil layer
was separated, while the aqueous layer was extracted with
toluene (3 ml ꢁ 50 ml). The solvent was evaporated, and the
residue was distilled under vacuum giving crude product
15 g (bp: 110–115 8C/1 mmHg), Recrystallization from
acetone and benzene gave pure 3 (14.5 g, yield 67%).
Melting point was 78–80 8C.
The thermal behavior of the polyimide 7 were investi-
gated by differential scanning calorimetry (DSC) and ther-
mograllimertic analysis (TG). Measurements were carried
out in nitrogen atmosphere at a flow rate of 100 ml/min and
at a heating rate of 10 8C/min. The results are summarized in
Tables 3 and 4.
The DSC analyses (shown in Fig. 1) show that the
polyimide 7d and 7d0 obtained from the reaction of
6-FDAwith 1,3-phenylenediamine have lower Tg comparing
with 7c and 7c0 [6]. While the Tg of 7d is lower than that of
7d0 which was prepared in DMF. Also the results of Tm
which were listed in Table 3 shown the substitute groups of
amines effect the Tm of polyimide was higher Tg using
aromatic amine in stead of alkyl amine.
FT-IR (nmax, cmÀ1): 3020 (w, Ar-H), 2950 (m, C–H), 1497
(s, Ar-H), 1180 (vs, C–F), 1120 (vs, C–F), 960 (s); 1H NMR
d: 3.3 (s, 12H, 4CH3), 8.0 (s, 6H, 6Ar-H); 19F NMR d: À64.0
(s, 6F, 2CF3); MS m/z (ion, %): 360 (Mþ, 100), 345
(Mþ À CH3, 10), 341 (Mþ À F, 9), 291 (Mþ À CF3, 98),
276 (Mþ À CF3–CH3, 56), 207 (Mþ À 2CF3–CH3, 27).
From Table 4, it is clear that all these polyimide 7 show
good thermal stability and rigidity. Among those, 7c is the
best one. It should be contributed to the rigidity of the
aromatic ring and molecular symmetry.
4.1.2. Preparation of 4,40-(hexafluoroisopropylidene)-
bis-(o-xylene) (3) from hexafluoropropene oxide
Hexafluoropropene oxide (HFPO) (17 g, 0.1 mol),
o-xylene (17 g, 0.16 mol) and AlCl3 (4 g, 0.03 mol) were
charged into a 100 ml autoclave. The mixture was heated to
110 8C for 16 h. After cooling, HFPO (8 g) was recovered. A
mixture of Ar2CCF3(CF2Cl), Ar2C(CF3)2 and Ar2C(CF2Cl)2
was obtained, and the overall yield was 25%. The ratio of
CF3 and CF2Cl was 2:5 (according to the 19F NMR).
3. Conclusion
We have successfully prepared the polyimide 7 as the con-
densation polymerization of 6-FDAwith diamines in different
solvents. The spectral, thermal and viscosity analyses of these
polyimide demonstrated that they have high thermal stability
with lower glass transition temperature. The polyimide 7c0,
preparedfrom 6-FDAwith1,4-phenylenediamineinDMFhas
the best thermal stability and higher molecular weight. The
electric properties of 7 are under further investigation.
4.2. Preparation of 4,40-(hexafluoroisopropylidene)-
bis-(phthalic acid) (4)
In a 250 ml flask, 4,40-(hexafluoroisopropylidene)-bis-
(o-xylene) (3) (7.2 g, 0.02 mol) was dissolved in a mixture
of pyridine (100 g) and water (50 ml) at 100 8C. The solu-
tion was stirred and refluxed gently for 4 h. During this
time potassium permanganate (15 g) was added carefully.
The hot solution was filtered and the cake of manganese
oxide was washed with a hot solution of pyridine (32 g) and
water (8 ml). The filtrate was evaporated to 40 ml, and
then added to sodium hydroxide (10 g, 0.25 mol) in
110 ml water. The solution was boiled in a 200 ml flask,
potassium permanganate (20 g) was added over 1 h. After
refluxing another 1 h, the solution was cooled somewhat and
the excess potassium permanganate was destroyed by cau-
tious addition of ethyl alcohol (2 ml). The mixture was
filtered, and the cake washed with hot water (40 ml). The
filtrate was evaporated until no residue pyridine can be detec-
ted and then acidified to pH ¼ 1 with hydrochloric acid.
The solution was evaporated to dryness and extracted with
acetone (3 ml ꢁ 50 ml). The acetone solution was evaporated
giving 6.3 g crude product (4), yield 66%.
4. Experimental
1H and 19F NMR spectra were recorded in CDCl3 on a
Bruker DRX-300 spectrometer operating at 282 MHz for
1
19F NMR (internal standard CFCl3) and 300 MHz for H
NMR (internal standard TMS). Highfield shifts from TMS
and CCl3F are negative. IR spectra were obtained with a
Perkin-Elmer 983G spectrophotometer using KBr disks.
Lower resolution mass spectra was obtained on a Finnigan
GC–MS 4021. Elemental analyses were undertaken by the
analysis department of Shanghai Institute of Organic Chem-
istry. Thermal analyses were performed on a SETARAM.
DSC-92 instrument. Intrinsic viscosity was measured on an
Ubbelodhe viscosimeter. Melting points were measured in a
melting point apparatus and reported uncorrected.