I DWIGHT E. PEERMAN, WESLEY TOLBERG, and DON E.

FLOYD
General Mills Research Laboratories, Minneapolis, Minn.

Reaction of Polyamide Resins and Epoxy Resins

Infrared analysis gives new information on curing con-
ditions in thermoset compositions used industrially

A m v o - c o n t a i n i n g polyamides, which such as plastics tooling (5, 6, 8, 9, 77). This article concerns a study of the
are essentially thermoplastic resins, may The reaction between amines and rates and degree of reaction of amino
be converted to thermosetting substances epoxy resins is rather complex (2, 3, 70). polyamide resins and epoxy resins.
by reaction with a number of materials, The complexity increases when poly- For comparison, data on commonly
including the epoxy resins. T h e amino meric coreactants such as the amino used short-chain amine curing agents
polyamide resins have interspersed along polyamides are utilized in place of simple are included.
the molecule, amino groups which may amines to obtain cross-linked structure. Thereactionofepoxy resins with amino
be primary, secondary, or tertiary and I n addition to the primary, secondary, polyamides is only mildly exothermic
which are free to react with, or catalyze and tertiary amine group in the poly- in nature, compared with that of epoxy
the curing of, epoxy resins (73). The amide, there may be heterocyclic nitro- resins with aliphatic short-chain amines.
thermoset compositions now find indus- gen-containing terminal groups such This suggests a slower rate of reaction.
trial usage as adhesives, coatings, and as the imidazolines or tetrahydropyrim- T h e short-chain amines generally harden
potting resins, and other applications, idines (7). epoxy resins rapidly, while polyamides

.
,

TIME IN MINUTES TIME - HOURS

Figure 1 . Relation of time to disappearance of epoxy Figure 2. Relation of time to disappearance of epoxy a t
at various temperatures 75" F.

VOL. 49, NO. 7 JULY 1957 1091

 require longer curing times a t room temperature and viscosity play a n im-
temperature. The amino polyamides portant part in the rearrangement of
probably react with epoxy resins more the molecules is also obvious. Viscosity
slowly because the high initial viscosity reduction a t elevated temperature pro-
slows down the reaction. Spectroscopic motes mobility and thus increases the
studies of epoxy resins in Great Britain rate of cross linking. Thus, although
( 4 ) have shown that ethylenediamine increase of curing temperature tends to
reacts more fully in a given time a t reduce the viscosity of the system and
room temperature than the amino poly- allow more complete orientation of the
amide. resin matrix, it also accelerates gelation
Therefore, the relation of residual and promotes immobility. The latter
epoxy content to cure conditions as eventually is predominant and a t this
well as to the physical properties of the stage the polymer has little further
thermoset reaction product was studied. chance for reaction. The thermoset re-
Obtaining accurate analytical data action product can thus become im-
is complicated by the fact that a t a cer- mobilized before all of the epoxy groups
tain stage in the reaction of amine curing are used up.
agent with epoxy resin, the viscous res- When a certain percentage of the
ins gel to a solid matrix. This solid amine groups available for reaction
is no longer soluble in organic solvents have reacted with epoxy groups, the
and cannot be melted. T h e apparent mixture becomes a solid and is ex-
solution to the problem was the use of tensively cross-linked. This inhibits the
infrared spectroscopic analysis. T h e in- reaction of the remaining amino groups,
frared spectroscopic techniques employed either in the polymeric curing agent or
were very satisfactory. despite consider- in the short-chain ethylenediamine. As
able mechanical difficulties, Prior in- the polymeric curing agent is more
frared work on products of this nature viscous than ethylenediamine, for in-
\vas helpful in developing the techniques stance, and reacts more sloiz-ly, it follows
used ( 7 , 72). that the optimum tempcrature for reac-
A new concept of optimum curing tion must be higher. If the temperature
conditions and of the interrelationship is too high, however, the matrix be-
of viscosity, temperature, and reaction comes solid or nonmobile before all of
rate was developed from data provided the epoxy groups present are able to
by the infrared analysis presented here. react fully. This finding is shown
T h e concept of molecular orientation graphically in Figures 1 to 3 and in
figure 3. Disappearance of
is postulated to play a part in the degree Table I. I n the work reported here,
epoxy at 200" F.
Vertical scale displaced of reaction between amino polyamides unless otherwise indicated, blends of
and epoxy resins which takes place 35 parts of Versamid 125 (trade name
under various curing conditions. T h a t of thermoplastic resin sold by General

100

80
I-
z
w
!-
z
0
0
>. 60
X
g
w
-I
4
f
40
0
LL
0

fl

20

, I I I
Figure 4. Infrared spectra of Versa- I 2 3 4
mid 125(35)/AraIdite 601 O(65) be- TIME IN HOURS
fore and after 10 minutes at 300" F.
Vertical scale displaced Figure 5. Rate of cure of Araldite 6010 at 300' F.

1092 INDUSTRIAL AND ENGINEERING CHEMISTRY

 EPOXY R E S I N S

lo0l
80

I
20
1
40
1

- MINUTES
60
I
80
1
100
I
120
'O0I
80

5 10 I5 20 25 30 35

Figure 6.
TIME
Heat distortion temperature as a function of
RESIDUAL EPOXY CONTENT % -
Figure 7. Effect of residual epoxy content on heat dis-
cure time
tortion temperature

Mills, Inc.) and 65 parts of Araldite scope of the present work to determine The optimum ratio of amino poly-
6010 (Ciba Co.) were used. these optimum temperatures. amide to epoxy resin was determined by
I t is concluded that if the optimum The disappearance of epoxy with other extensive physical testing, including
temperature of reaction is selected, the typical amine curing agents has been such tests as heat distortion temperature,
competing factors of decrease in viscosity discussed (7) ; estimation of residual tensile, flexural, and compressive
due to heat and increase in immobility epoxy is in essential agreement with the strength, hardness, and impact resist-
due to reaction can be balanced suc- findings in the present study. ance. Heat distortion temperature has
cessfully, so that nearly all of the epoxy Disappearance of oxirane oxygen con- been shown to be a sensitive indication
groups in the epoxy resins react with tinued even when insufficient amounts of degree of cure.
amine groups in the amino polyamide. of coreactant were present. I n demon- T h e present work shows clearly that
Some of the more commonly used strating this phenomenon, three ratios heat distortion temperature is related
amine-type curing agents are not cured of amino polyamide to epoxy resin very closely to the disappearance of
in an optimum manner a t 300' F., a were chosen. When excess polyamide epoxy groups in the polymer (Figures
frequently suggested curing temperature. was used in the blend, the rate of disap- 7 and 8). T h e heat distortion is a
Figures 4 and 5 show that at 300' F. the pearance of epoxy was approximately measure of the resistance to flexural
blend of amino polyamide and epoxy the same as when the optimum ratio
100
resin contains essentially no residual of amino polyamide to epoxy was used.
epoxy groups after 10-minute cure, but (The optimum ratio of amino polyamide
appreciably epoxy content is left in epoxy to epoxy resins is defined arbitrarily as
and amine hardener systems under the the blend having the highest heat dis-
same conditions. Although the short- tdrtion temperature.) Thus, a t 300' F.,
chain amine-cured epoxies have good both of these ratios left essentially no
physical properties when cured a t 300°F., oxirane oxygen content at the end of a
c
this is obviously not the temperature 10-minute heat-up period, during which
2
*(
at which all epoxy content disappears. time the temperature attained 300" F. B D

Perhaps a more suitable curing tem- When a deficiency of polyamide in re- x

2.

perature would give improved properties lation to epoxy was used, the disap- 2
50
in these systems. I t was not within the pearance of epoxy assumed an entirely -1

different 'course. Epoxy initially dis- r"

appeared rapidly, but when the avail- % 40

Table 1. Disappearance of Epoxy able amine had been exhausted, there b

r(
Groups in Blends on Curing at 75' F. was a continuing dissapparance of 30
epoxy, which was attributed to either
Time % of degradation or catalyzed eelf-polymer-
Interval, Origmal 20
Hours Epoxy Content ization. T h e disappearance of epoxy
0 100
...
with the various ratios at constant tem-
10
2 92.5 perature is shown in Figure 6. Epoxy
26.5 45.0 continues to disappear up to and through
31 43.9 12 hours, even though the amino groups I
95 2 4 6 8 1 0 1 2
41.7
124 41.2 of the polyamide have been used u p TIME IH HOURS

165 40.8 quickly in the initial portion of the reac- Figure 8. Effect of polyamide-epoxy
tion. ratio on disappearance of epoxy

VOL. 49, NO. 7 ' JULY 1957 1093

 a t 75’ F. Full cure does not take PLUNGER
place even after a postcure a t 300’ F.
This finding substantiates the concepr
of formation of a solid matrix a t other
than optimum temperatures, which
greatly hinders complete disappearance
of epoxy in the system.
The oxirane oxygen content of the
samples studied was determined by
observing an absorption peak in their
infrared spectra a t about 10.9 microns. R
The samples were prepared as thin
1 1 7 0 HOURS films (0.002 inch) between potassium
chloride plates. The spectra were re-
corded a t essentially zero time and
after approximate intervals, during which
23 HOURS the samples were cured a t various tem-
peratures.
Initially, the most obvious approach
in preparing resin samples for de-
termination of their infrared spectra
DIE
was to form a thin film of known thickness
between rock salt plates separated by Die for potassium chloride plates
a standard spacer. O n curing, the
resin blend creates a strong bond be-
tween the plates and prevents recovery
2 HOURS of the rock salt plates for re-use. Diffi-
culty was encountered in preparing than absorbance of the epoxy band were
suitable spacers. required. As the same sample was ob-
An alternative to this method was served throughout cure, errors in manip-
developed. I t was relatively simple to ulation, sample thickness, and time of
prepare clear plates, 1 inch in diameter cure were avoided. At least two sam-
0 TIME and about ”18 inch thick, from powdered ples were studied under each set of con-
potassium chloride. These were made ditions and agreement was obtained in
Figure 9. Disappearance of epoxy
in a pellet die under 50-ton pressure. each case.
a t 75”F.
Vertical scale displaced O n e of each pair of plates was made
with a ram which had a 0.001 inch rise Literature Cited
bending under the influence of elevated across the center ”4 inch of its face.
This formed a 0.002-inch depression (1) Dannenberg, H., Harp, W. K.,
temperature. Other physical measure- Anal. Chem. 28,86 (1956).
ments, such as hardness, ultimate flex- between the plates, which allowed resin (2) Greenlee, S. 0. (to Devoe-Raynolds),
ural strength, and compressive strength, thickness to be maintained even under U. S. Patent 2,585,115 (1952).
are less closely related to the degree of pressure. Plates made with this ram (3) Zbid., 2,589,245.
had what amounted to a built-in spacer. (4) O’Neil, L.A., Cole, C. P., “Chemical
reaction. Many partly cured specimens and Spectroscopic Studies of Epoxy
show considerable physical strength. T h e resin blend was placed between Resin Reactions in the Surface
The relationships of heat distortion tem- potassium chloride plates and installed Coating Field,” Birmingham Paint
in a special holder which firmly com- Club, Birmingham, England, 1956.
perature, hardness, and flexural strength (5) Paint, Oil C3 Chem. Rev. 113 ( 2 3 ) , 15
in cured blends of amino polyamides and pressed the sandwich so that sample
(1950).
epoxy resins are shown in Table 11. thickness would not change on heating. (6) Peerman, D. E., Floyd, D. E.,
Heat distortion values are critical to the As the samples were of constant thick- Mitchell, W. S.. Plastics Technol. 2,
elements of the study and determinations ness, the consecutive readings of ab- 25 (January 1956).
sorbance were compared directly with (7) Peerman, D. E., Tolberg, W, E.,
are repeated no less than four times. Wittcoff, H., J. Am. Chem. Soc.
Barcol hardness is routinely measured the reading a t zero time to obtain per 76, 6085 (1954).
ten times. Flexural strength data are cent oxirane oxygen. The sandwich (8) Renfrew, M. M., Wittcoff, H., FIoyd,
obtained in duplicate as they are not methods thus permitted determination D. E., Glaser, D. W., IND.ENC.
of epoxy content on samples too soft for CHEW46,2226 (1954).
critical. (9) Riley, M. W., “Plastics Tooling,”
The infrared curves of blends cured conventional potassium chloride plate Reinhold, New York, 1955.
a t room temperature (75’ F.) (Figure 9) solid state dispersions and too insoluble (10) Shechter, Leon, Wynstra, John,
show the epoxy to react only partially for solution spectra. No data other Kurkjy, R., IND. END. CHEM.
48,74 (1956).
(11) Silver, I., Atkinson, H. B., Jr.,
Modern Plastics 28 ( 3 ) , 113 (1950).
(12) U. S. Dept. Commerce, Office of
Table It. Heat Distortion Values for Blends under Various Curing Conditions Technical Services, Bull. C. B.
Temp., Time, Barcol Heat Distortion, Flexural 1111438 (1954).
O F. Min. Hardness ’F. Strength (13) Wittcoff, H., Renfew, M. M. (to
General Mills, Inc.), U. S. Patent
400 5 79 174 13,600 2,705,223 (1955).
400 60 80 176 14,900
300 10 77 171 14,600 RECEIVED
for review October 19, 1956
300 90 86 187 14,800 ACCEPTED February 13, 1957
200 10 55 131 13,400
200 30 69 156 11,500 Division of Paint, Plastics, and Printing
200 60 68 154 12,000 Ink Chemistry, Symposium on Epoxy
200 120 72 162 9,100 Resins, 130th Meeting, ACS, AtIantic
200 480 77 171 11,000 City, N. J., September 1956. Journal
Series 203, General Milk, Inc.

1094 INDUSTRIAL AND ENGINEERING CHEMISTRY