close

Вход

Забыли?

вход по аккаунту

?

Analysis of resorcinol-phenol-formaldehyde resins by NMR spectroscopy.

код для вставкиСкачать
Die Angewandte Makromolekulare Chemie 26 (1972) 171--176( N r . 379)
From the School of Chemical Sciences,Universityof East Anglia,Norwich,NOR 88C,
England (R. A. and A. H. H.), and the Research Department, CIBA-GEIGY (UK)
Ltd., Plastics Division,Duxford,CambridgeCB2 4 &A, England (B.P.S.)
Analysis of Resorcinol-Phenol-FormaldehydeResins
by NMR Spectroscopy
By R. ANDERSON,
A. H. HAINES,and B. P. STARK
(Eingegangenam 7. April 1972)
SUMMARY:
A method is reported, based on NMR spectroscopy, for approximate determination of the component ratios of resins formed by condensation of resorcinol, phenol
and formaldehyde.
ZUSAMMENFASSUNG:
Es wird eine NMR-Mcthode zur Bestimmung der ungefiihren Verhiiltnisse zwischen aus Phenolkernen, Resorcinkernen und Fornialdehyd gebildeten Gruppen
in Kondensationsprodukten aus Phen 01, Resorcin und Formaldehyd beschrieben.
Resins formed by the co-condensation of resorcinol, phenol and formaldehyde
are of considerable industrial importance as weatherproof wood glues, and also
as adhesives for bonding asbestos-based boards, various plastics, and other
materials. Their advantages include chemical stability, simplicity in use, and
the ability to cure a t room temperature. At the moment, very few methods
exist for the analysis of R P F and R F resins, and a convenient method for
determination of their component ratios is desirable. The present paper concerns
an investigation into the use of NMR spectroscopy for this purpose, using an
approach which is in some respects an elaboration of that applied by WOODBREY
and co-workers to the analysis of phenol-formaldehyde resinsl. These workers
showed that the protons in different chemical environments in PF resins (or
their acetylated derivatives) could readily be distinguished by virtue of their
different chemical shifts in NMR spectra. From consideration of the integrals
for protons in the different functional groups, important structural parameters
of PF resins could be determined (e.g. number average molecular weights and
average numbers of each functional group per aromatic ring).
171
R. ANDERSON,
A. H. HAINES,and B. P. STARK
Results and Discussion
The technique used in the present work involves acetylation of an RPF (or
RF) resin, and consideration of the numbers of aromatic protons removed on
reaction of resorcinol or phenol units with formaldehyde to form inter-nuclear
links (e.g. methylene bridges) or functional groups (e.g. hydroxymethyl groups).
Consider a resin formed by the reaction of 1 mole of resorcinol with x moles
of phenol and y moles of formaldehyde. The formaldehyde originally present
will be combined in links of the types: Ar-CHz-0(methyleneoxa, mo*:
Ar -CHzOH, Ar -CHzOCHzOH, Ar -CHzOCHz -), Ar -CHz -Ar (methylene
bridge, mb), or Ar-CH~-O-CHZOH
(hemiformal, hf). On acetylation, two
types of acetate groups, Ar-OCOCH3 (phenol acetate, pac) and Ar-CH2OCOCH3 (hydroxymethyl acetate, rnac) will be introduced**. Aromatic protons
(ar) will remain unaltered on acetylation of the resin.
By examination of the spectra of the dried acetylated RPF resins, and from
consideration of earlier work on PF resinsl, the integrals (A) for the various
types of protons in different environments can be measured in the following
regions of the NMR spectrum: A,, 7.67-6.58 ppm; Ahf 5.28-5.17 ppm; Am,
5 . 1 4 . 2 5 ppm; Amb 4.03-3.42 ppm; Apac 2.33-2.13 ppm; Am,, 2.07-1.92 ppm.
The following two relationships then hold :
Total aromatic hydrogen in original RPF mixture
-
Total phenol acetate hydrogen in acetylated product
4 4-5X
Aar C 4 Am0 4-Amb -Apac
and
Total formaldehyde-derived hydrogen
-=-Total phenol acetate hydrogen
6
+ 3x
2Y
6
+ 3x
- Ahf
+ Amo + Amb
Apac
(2)
Thus, x and y may be calculated, if the relevant integrals are measured from
the NMR spectrum.
The peaks due to the methyl protons of phenol acetate and hydroxymethyl
acetate groups frequently overlap to such an extent that an accurate value of
Apaccannot be measured directly. For resins not containing benzyl hemiformal
acetate species (A~CHZOCHZOCOCH~),
a value for Apac may be obtained by the
following procedure : (a) the integral for the methylene protons of the hydroxy-
*
**
The abbreviated designations mo, mb, hf, pac, rnac, and ar are introduced to
identify conveniently the various integrals measured on the NMR spectrum.
For the purposes of the calculation, the group Ar - CHzO - C H 2 0 -COCHs (hemiformal acetate) is included with A,,, since the methyl protons of the t w o
groups have similar chemical shifts. No hemiformals were in fact detected in
this work.
172
Phenol Resins, N M R Spectroscopy
methyl acetate group, Ar--CH~-OCOCK3, is measured, (b) the integral value
is multiplied by 312, and (c) the resultant figure is subtracted from the value
of (Apac Amac). The NMR peaks for the methylene protons (5.14.99 ppm)
are sufficiently well separated from those of the other mo-type protons to permit
accurate integration.
The feasibility of using an NMR technique based on the above reasoning was
tested, by analysing a number of R P F and R F resins ; some results are summarised in Table 1. R P F resins were prepared by two different procedures : direct
acid-catalysed reaction (Method “X”) and two-stage base-catalysed condensation (Method “Y”) (see Experimental). The R F resin was made by a process
employing an acid catalyst.
I n order for this test of the NMR analysis to succeed, it is obviously necessary
that there should be no loss of materials prior t o obtaining the spectra. The
analysis of each resin as its acetylated derivative was therefore fist attempted
without preliminary drying, in order that the possible removal of labile formaldehyde or of phenols might be avoided. I n view of the high reactivity of
resorcinol, it was not expected that large amounts of labile formaldehyde (in
the form of either hemiformals or polyoxymethylene oligomers) would be
present; the spectra in fact were found to contain no peaks attributable to
acetate derivatives of benzylic hemiformals (Ar-CHZO-CH20H),
benzylterminated polyformals (Ar- (CHzO),CHzOH ; x > 2), or polyoxymethylene
oligomers (HO(CHzO),H), nor were NMR peaks characteristic of benzyl ether
bridges (Ar-CH~-O-CH~-Ar)
observed. It was presumed that treatment of
the undried resins with an excess of acetic anhydride and pyridine would yield
only water-soluble by-products, which would be removed during isolation of the
acetylated derivative. Large discrepancies between the measured and expected
ratios were however observed with resins which were not dried before acetylation (see Table 1) ; such errors were presumably due to difficulty in achieving
complete acetylation of the resin in the hydroxylic medium. When the ratios
were normalised with respect t o resorcinol (the least volatile component), the
measured phenol and formaldehyde contents were invariably lower than in the
initial mixtures.
Considerably better results were obtained when the resins were dried before
acetylation (seeTable 1). WOODBREY
and co-workers1have shownthat the freezedrying of heat-reactive resins (resols) derived from phenol and formaldehyde
leads to removal of formaldehyde from benzyl-terminated polyformals and from
polyoxymethylene oligomers, and also that free phenol is removed unless the
resin is basic. Since formal groups did not appear to be present in any of our
resins, it appeared reasonable t o assume that the drying of these would not alter
their formaldehyde content. Moreover, in the case of resins produced by method
+
173
Initial molar RPF ratios
(ascharged)
Method of
Preparation
are normalised with respect to resorcinol
Standard deviations are given t o two decimal places.
b Acid-catalysed condensation (see Experimental).
c Base-catslysed condensation (see Experimental).
d Acid-catalysed condensation.
e Dried for 1hr. at 0.5 mm.
a Ratios
Entry
No.
Undried
Ratios from NMR analysisa
Table 1. NMR Analysis of component ratios in resins derived from resorcinol, phenol and formaldehyde.
Dried
Ew
u1
?J
Pj
P
P
Eo
Phenol Resins, N M R Spectroscopy
“X” (seeExperimental, and entries 1 , 2 and 3 inTablc l),the loss of free phenol
on drying for ca. 2 hrs. a t 0.5 mm does not appear t o be a significant source of
error. Since resins produced by method “Y” (see Experimental) are basic, they
should not lose phenolic components on drying. I n view of the greater possibility
of formal type linkages being present in the latter type of resin (entries 4 , 5 and
6 in Tab. 1) -but presumably in amounts less than that detectable by NMR
- the drying period for them was reduced to 1 hr a t 0.5 mm.
If it is assumed that resorcinol is determined exactly by this NMR technique,
then the errors in the series of phenol and formaldehyde determinations made on
dried samples vary from 0 to 21 yoof the values expected from the initial charge
ratios. The analysis of RPFresins containing hydroxymethyl groups (e.g. entries
4, 5, 6 in Table 1) is inherently open to larger errors than the analysis of those
without them (entries 1, 2, 3), since with the former type of resin APac must
generally be determined indirectly, as described previously. Although the
accuracy of the analysis may not be high in absolute terms, information within
these limits is still of considerable practical value in many instances. The relative
rapidity with which useful information may be obtained on this hitherto
intractable problem illustrates further the potential of NMR analysis in the
field of resins derived from formaldehyde and phenolic compounds.
Experimental
The RPF resins were prepared under acidic or basic conditions, as indicated
below, and were then acetylated before NMR analysis. NMR spectra were run at
3OoC in CDC13, using a Perkin-Elmer R 12 instrument, with tetramethylsilane as
internal standard. Chemical shift data are reported here as 6 values. The results in
Table 1 were calculated on an ICL 1905 Series E computer, using EL program based
on equations (1)and (2). Six values of each integral were measured on the spectrum.
These were entered into the appropriate data section of the computer program,
which contained a subroutine to evaluate the mean value and standard deviation of
each of these sets of integrals. The latter values were then used in the main program,
which calculated the component ratliosof the resin and, by a usual statistical procedure2, thoir standard deviations.
Preparation of Resins under Acidic Catalysis (Method “ X ” )
The following procedure was typical. Water (20 ml) and phenol (21.96 g, 0.233
mole) were heated with stirring to 65OC, and resorcinol(36.7 g, 0.333 mole) waa then
added, the temperature of the mixture being maintained constant during this time.
The solution was acidified t o pH 2.5 by the dropwise addition of 98% formic acid,
and was then heated at 100°C for 30 min. The temperature of the mixture was allowed t o fall and was maintained at 90 “C, whilst 37.4% formalin (24.0 g, 0.299 mole of
formaldehyde)was added during 160min. After a further 2 hrs. at 95”C,the reaction
mixture was cooled, and brought to pH 7 by the addition of 20% sodium hydroxide
175
R. ANDERSON,
A. H. HAINES,
and B. P. STARK
solution. The resin was either acetylated immediately or dried in vacuum (0.5 mrn)
for 2.5 hrs., over calcium chloride.
Further resins were prepared similarly, w-ith molar R P F ratios of 1/0.?/0.7 and
1/0.5/0.7.
Preparation of Resins under Basic Catalysis (Method ‘‘ Y”)
I n a typical experimeiit, a mixture of phenol (28.25g, 0.30mole), 37.40/0 formalin
(28.1 g, 0.35 mole of formaldehyde) and 50% sodium hydroxide solution (0.27 g)
was heated to 80°C with stirring. An exothermic reaction raised the temperature of
the mixture to its boiling point, and reflux was maintained by applying external
heating for 15 rnin., a,fter which time the temperature was maintained a t 90°C for
1 hr. Resorcinol (11g, 0.0999 mole) and 50% sodium hydroxide solution (0.82 g)
were then added, and the mixture was heated to boiling under reflux and with stirring for 30 min. The resin was then kept at 90°C for 30 min., and 80°C for 20 min.,
after which time it was allowed to cool, and water (21.5 ml) followed by 50% sodium
hydroxide (1.68g) were added. The resin was either acetylated immediately or
dried for 1 hr. over calcium chloride in vacuum (0.5 mm).
Further resins were prepared similarly, with R P F ratios of 1/2.2/2.8 and l/l.66/
2.33.
Acet ylation Procedure
The dried resins (0.3 g in each case) were acetylated by the procedure of HIGGINand co-workers3, using dichloromethanc in the isolation procedure. When
an undried resin was acetylated, double quantities of acetic anhydride (4.0 ml) and
pyridine (1.0 ml) were used.
BOTTOM
J. C. WOODBREY,
H. P. HIGGINBOTTOM,
and H. M. CULBERTSON,J. Polymer Sci.
A 3 (1965) 1079.
2 D. C. BAIRD,An Introduction to Measurement Theory and Experiment Design,
Prentice Hall, Englewood Cliffs, 1964.
3 H. P. HIGGINBOTTOM,
H. M. CULBERTSON,
and J. C. WOODBREY,
Anal. Chem. 37
(1965) 1021.
1
176
Документ
Категория
Без категории
Просмотров
0
Размер файла
311 Кб
Теги
spectroscopy, nmr, resins, analysis, formaldehyde, phenols, resorcinol
1/--страниц
Пожаловаться на содержимое документа