Urea-Formaldehyde FilmForming Compositions Study of Structure and Properties
The unusual light color, low cost, and heat-hardening characteristics of ureaformaldehyde condensation products have stimulated a great deal of research on the application of these resins for surface coatings. The progress towards making a satisfactory coating composition from urea and formaldehyde alone appears to be effectively barred because of (a)the lack of solubility in the commonly used inexpensive solvents, ( b ) the lack of stability, and (c) the lack of compatibility with resins and drying oils. The purpose of this paper is to describe a product which is useful to the coatings industry and to present the theoretical background upon which this product is based.
T.S. HODGINS AND A. G. HOVEY Reichhold Chemicals, Inc. ,1 Detroit, Mich.
A
BlZIEF survey of the chemistry of the urea-formaldehyde reactions as made by Ellis (24) shows that many investigators were attracted to this field prior to the work of John (32) on urea-formaldehyde condensation products. Shortly after John’s announcement, numerous investigators described many urea-formaldehyde condensation products which were intended either as molding binders or adhesive materials. John’s product was prepared with an excess of formaldehyde and without a condensing agent, although it is now generally considered that the formic acid present in his formaldehyde acted as a Condensation agent (&). I t was proposed for use as an adhesive, a molding binder, and a thermosetting impregnant for fabrics. The introduction of organic solvents into the reaction mixture for the purpose of dehydrating the resinous condensation products was described by Ramstetter (38) and Crump ( 1 8 ) . Another step towards the development of urea-formaldehyde 1
Foimerly Beck, Koller & Company, Inc.
resins as coating compositions was a discovery by Lauter (33) that urea and formaldehyde would react in the presence of specific quantities of glycerol or other polyhydric alcohols to produce resins which are soluble in alcohols and other nonaqueous solvents. Other methods of dehydrating an aqueous condensate in the presence of organic solvents were described by Ripper (41) and by Luther, Pungs, Griessbach, and Heuck (36). Hill and Walker (29) reported the condensation of urea-formaldehyde intermediates in the presence of oil-modified alkyd resins and monohydric alcohols. The authors (31) described the co-condensation in monohydric alcohols of urea, formaldehyde, and a modifier consisting of polybasic acid-polyhydric alcohol intermediate products employing an excess of the polyhydric alcohol. Cheetham and Pearce (16) wrote regarding reaction products of urea and formaldehyde as finishing materials.
General Reactions REACTIONSIN AQUEOUSMEDIA. In an aqueous medium numerous investigators seem to be more or less agreed on the general principles underlying the reactions of urea and formaldehyde. In general, the resulting condensation product is dependent upon (a) the catalyst employed, including the pH of the solution as a catalytic factor, ( b ) the molecular ratio of the reacting components, and (c) the time and temperature of reaction. With one mole of urea and one mole of
RESIN KETTLE 1021
INDUSTRIAL AND ENGINEERING CHEMISTRY
1022
formaldehyde in the presence of hydroxyl ions, monomethylolurea is formed (23):
xHcHzoH
NHz
c-0 + L o -
1
=o
When an excess of formaldehyde (i. e., one mole of urea and two moles of formaldehyde) is employed, dimethylolurea is formed (23):
rHz cl
NHCHzOH
2CHz0
+
1
=O
GO+
(2)
I NHCHzOH
NHz
The hypothesis is advanced by Redfarn (39) that resin formation takes place according to the polymerizationof dimethylene urea as a saturated chain:
,-.
( 1)
NHz (monomethylolurea)
hTH2
(dimethylolurea)
This appears quite improbable because such structure indicates linear polymers which, according to resin theory, would be a heat-nonconvertible resin. The case, however, is that urea-formaldehyde polymers definitely behave like three-dimensional units because irreversible gel structure is readily obtained on continued heating. Another reaction mechanism is suggested by Ellis (266) whereby two urea molecules are reacted with three formaldehyde molecules to form one mole of monomethylolurea and one mole of dimethylolurea, which, in turn, are assumed to form methylolcarbamyl-4-ketohexahydrotriazine: H
I n the presence of dilute acids, methyleneurea, a white amorphous insoluble precipitate, is formed (34):
CHzO
+
FHz + , =O
VOL. 30, NO. 9
H N
acid
----f
/\
O=C,
/
CHZ
+ HzO
(3)
‘Hd(methyleneurea)
i6HZ
Two moles of this material could condense to form CO-NH
or, according to Holzer (30): CO-NH
“2
CHzO
/
NHz
‘CO.NH. CHA
I
+
I
N=CHz
\
\
\
1
N=CH2
REACTION MECHANISM IN AQUEOUS MEDIA. The mechanism of reaction has had many interpreters. Kotable among them are Goldschmidt (26), Dixon (ZI), Staudinger (46), Scheibler and co-workers (43),Walter and co-workers (47,48), and de Chesne (17‘). There appear to be two possible mechanisms by which the final end product from the reaction of one mole of urea and two moles of formaldehyde is obtained: (a) condensation through. the dimethylolurea intermediate and (b) polymerization through the dimethyleneurea compound : NkCHz d=O 2H20 f-
(7)
/CH2 CHz-N
+ d=O + acid +A=O “2
\
1 1
+ 2CH20
NHCHzOH &=O &HcH,OH / (-HzO) /
(4)
The unit polymer is: CO-NH H-
-N
Owing to the fact that the molecular proportions are so chosen (two moles of urea to three moles of formaldehyde with the subsequent formation of one mole of dimethylolurea and one mole of monomethylolurea), there is a possibility that the methylolcarbamyl-4-ketohexahydrotriazinecould form. Yet for the same reason that Redfarn’s hypothesis seems improbable (i. e., because linear polymers are not heat convertible), these structures shown in Equations 7 and 8 do not appear satisfactory as a polymer unit because of the lack of provision for cross linkages which are necessary to account for gelation. Experimentally the writers have found that the resin does gel. Gelation is not accounted for unless it is shown that cross linkages occur by the reaction of additional molecules to tie in a t some other place between the linear polymers. If the reacting molecules were all dimethylolurea (possible when starting with one mole of urea and two moles of formaldehyde) instead of an equal number of molecules of mono- and dimethylolurea (as was the case when Ellis started with two moles of urea and three moles of formaldehyde), the chance probability of reacting molecules in the three dimensional space appears to be highly in favor of the forma-
SEPTEMBER, 1938
INDUSTRIAL AND ENGINEERING CHEMISTRY
1023
COMPLEXPOLYHYDRIC ALCOHOLS.The polyhydric alcohols used were ethyleneglycol and glycerol, and were extended to include the reaction product of one mole of phthalic anhydride and two moles of ethyleneglycol which, for theoretical considerations, will be written as follows :
b While such compounds have never been isolated, it is obvious that compounds resulting from the reaction of one mole of phthalic anhydride and two moles of ethyleneglycol are substantially complex dihydric alcohols. Similarly, other complex polyhydric alcohols may be prepared from three moles of phthalic anhydride and four moles of glycerol and would be of varying chain length; but in general they might be expected to be similar to reaction 10: CHzOH I
0
~HOH
A-k-0--LH2
ALKYDMODIFIERKETTLE
tion of the structure shown in Equation 4. Since monomethylolurea is tetrafunctional (three reactive hydrogen atoms and one reactive hydroxyl group), it appears that, even if monomethylolurea alone were involved without any molecules of dimethylolurea, it should still be possible to form cross linkages between linear structures and account for heat convertibility. Some dimethylolurea is generally present with the monomethylolurea which greatly increases the chances of forming complex structures and expediting heat convertibility.
Urea-Formaldehyde Condensation with Mono- and Polyhydric Alcohols ACETALFORMATION FROM MOXO-AND POLYHYDRIC ALFor urea-formaldehyde resins to be useful to the paint and varnish industry, they must be compatible with the commonly used paint and varnish solvents, resins, and lacquers. Resins made in aqueous solvents do not have the property of compatibility with- paint and varnish materials. Urea-formaldehyde products condensed in the presence of certain monohydric alcohols and/or polyhydric alcohols are soluble in hydrocarbons. Since the properties of ureaformaldehyde resins are based upon the nature of the medium of reaction, whether aqueous or alcoholic, the authors believe that in either case the medium plays a part in the resinforming reaction-i. e., the basic reaction of formaldehyde with mono- and/qr polyhydric alcohols to form acetals and the subsequent reaction of the acetals with urea to form the unit polymer. Butyl alcohol was chosen as the monohydric alcohol, chiefly because of its ease of reaction and its property of forming constant-boiling mixtures with water. COHOLS.
Owing to the relatively low temperature (190' C.) and to the molar excess of glycerol with respect to phthalic anhydride, we should expect that very little reaction would take place a t the beta position of the hydroxyl groups of the glycerol. In the preparation of these two complex polyhydric alcohols (reactions 9 and lo), the acid number is used as a controlling factor for obtaining a repetition of the same end point. Since monohydric and polyhydric alcohols both form acetals readily with aldehydes (Table I), it is assumed that such complex polyhydric alcohols as have just been described (reactions 9 and 10) may form complex acetals. The acetals may, in turn, react with urea in the same manner as the simpler acetals. The industrial uses of these complex polyhydric alcohols as modifiers in the formation of urea-formaldehyde condensation products will be discussed later in this paper. FORMATION OF n-BuTYL FORMAL. I n the condensation of urea and formaldehyde in the presence of a monohydric alcohol, such as butyI alcohol, the process differs from that which occurs in aqueous media because of the initial reaction of the alcohol with the aldehyde.
INDUSTRIAL AND ENGINEERING CHEMISTRY
1024
+ 2HOCHzCH2(O.CHp*CHa.CHz*CHs)z CHa CHz CHa + HzO
CHzO
1
*
----f
*
(11)
(n-butyl formal) *Butyl formal is formed in the presence of an excess of butyl alcohol. If a limited amount of the alcohol is presenti. e., one mole of alcohol per mole of formaldehyde-the tendency is to form the hemiacetal: OH
CHlO
+ HOCHz CHa *CH,*CHs+CHs *
/ \
(12)
0 .CHz. CHz .CH2 * CHI
in-butyl hemiformal)
PROPERTIES OF ACETALS. Table I gives a summary of some of the more important compounds from alcohol-aldehyde reaction products which might serve as intermediates for reaction with urea to form resins compatible with surfacecoating materials. The general acetal reactions are: Acetal: RCHO
+ 2R10H +RCH(OR1)z + HzO
(13)
OH
Hemiacetal: RCHO
+ RIOH +RCH/
(14)
\OR,
REACTION OF UREAWITH ACETALS. Urea reacts with the initial alcohol-aldehyde compound (an acetal) to form the condensation unit on a theoretical basis as follows: NHz
d
--O
~
+ 2HzC(OCHz.CHz.CHz.CHs)z +
H
Z
H NCHzO CHz .CH .CHz * CHI
/ c=o
+ 2CHa.CHz. CHz.CHzOH
\
(15)
NCHzO CHZ * CHz * CHa * CHs H
The condensation may follow as in the aqueous medium (reaction 4) : HNCHz OR LI
H ~
O
H NCHzOR LI
O
C OR H ~ H NCH~OR
RO CHzN H
I c=o
ROCHA H
(16)
RO CHzNH
~
H
Z
1 mole 60
++ 260moles
- -
2 N = 28 = 33.3% 3C 36 42.8 0 = 16 = 19.1 4H = 4 = 4.8
- 100.0 84
A nitrogen analysis of the urea-formaldehyde condensation product reacted in the presence of water gives results somewhat higher than the theoretical (nitrogen = 35.2 per cent). Dixon (91) reports a nitrogen content of 32.8 per cent, which is in close agreement with the theoretical. Comparison of the theoretical and actual nitrogen analysis, together with the fact that the theoretical yield of resin was obtained, would seem to indicate that the unit polymer has the general structure given in reaction 17. If the reaction of one mole of urea and three moles of formaldehyde is carried out in the presence of butyl alcohol, Urea Aqueous formaldehyde Butyl alcohol
3'1
60 grams
243 74
-
a77
the yield of resin obtained is 153 grams. On a basis of 153 grams per mole of urea, the yield is 2.55 grams of resin for 1 gram mole of urea as compared to 1.40 grams of the resin obtained in the aqueous medium per gram mole of urea. This fact alone shows that butyl alcohol definitely enters into the reaction with urea and formaldehyde. After co-condensation of 40 grams of ethyleneglycolphthalate modifier (reaction 9) and 60 grams of urea-formaldehyde condensation product, the resulting resin showed a nitrogen content of 13.0 per cent by the Kjeldahl method. While it may be considered that urea-formaldehyde and ethyleneglycol-phthalate condensation products are present simply as a physical mixture, we believe by reason of analogy that the ethyleneglycol-phthalate modifier behaves as a complex dihydric alcohol and reacts with the urea-formaldehyde condensation product in the same manner as butyl alcohol has been assumed to react. Because of the complexity of the structure of the glycolphthalate modifier, its reaction products with the ureaformaldehyde condensation product, which is already a complex structure, would make the resulting condensation product exceedingly complicated. If one unit of the ureacondensation product and one unit of the ethyleneglycolphthalate modifier are joined according to the structure I
$=OH
-I~-cH~-o-cH~
I c=o I
cH2-0-+n 1
o
RO CHz"
NHCHzOH
+ 2CHz0
This unit of structure, according to theory, would have the following composition :
-N-CHZ-
+ n ROH
AQUEOUSAND ALCOHOLIC REACTIONPRODUCTS. When we react one mole of urea (60 grams) with two moles of 37 per cent formaldehyde (162 grams) in a closed system, we obtain 84 grams of resin in 222 grams of resin solution, or 37.9 per cent. This is equal to 84 grams of resin per mole of urea, a ratio of 1.40 to 1. From these data i t may be assumed that the following type of reaction takes place:
:x
VOL. 30, NO. 9
b=O
--t
I-CHrI =O
+ 2H20
I AHCH9OH N-CHZI 1 mole 120
++
(17)
1 mole unit 2 moles 84 36
there are many reactive points' left for further reaction. The nitrogen content of this hypothetical unit is 8.47 per cent, whereas the actual analysis showed 13.0 per cent. However, since the urea-formaldehyde polymer unit can react not only with the ethyleneglycol-phthalate unit but also with other units of its own kind, it i s quite possible to form compounds having a nitrogen content much higher than 8.47 per cent, with 33.3 per cent as a limiting value because this is the nitrogen content of the urea-formaldehyde unit polymer. On the other hand, there is always the possibility of a mechanical mixture to be considered. If the reaction of the urea-formaldehyde unit with the complex polyhydric alcohol unit should proceed farther than is shown in reaction 18,
INDUSTRIAL AND ENGINEERIXG CHEMISTRY
SEPTEMBER, 1938
TaBLE
Empiriaal Formula
I. ACETALSFORNED AI.
Structural Formula
1025
P.
B. P.
c.
Sp. Gr.
Literature Citations
c.
From Formaldehyde
....
CHz(0CzHs)z CHz(0.CHz.CH:CHz)z CHz(0. CHz.CHz.CH:CHz)z CHz(0 .CHz.CHz. CHz.CHa)z CHz[O ~CHz.CH(CHa)zlz CHz[O. CHz. CHz CH(CH3)zIz CHz[O.CHz.CH(CHa) .CHz.CHa]z CHz(OCsHi7)z
/OCHz CHz
....
....
-61.6
....
.... ....
....
87-88 138-139 175-177 180-181 164.3 207.3 205 360
....
74-75
97-8
198
0.851 (0' C.) 0.8948
....
....
0.825 (20' C.) 0.835 (20' C.) 0:846 (200 C.)
1.0534 (25O C . )
\
CHOH
'OCHz/
I
/°CH'CH20H CHz \OCHz HzC-CH
1
I
O.CHz.0
.HC-CHz l l O.CH2.O
From Acetaldehyde and Higher Aldehydes CHs CH (0 CH3)z CH3"2H(OH) .OCzHs CH3. C H ( 0 .CzHs)z C H s ' C H ( 0 . CsH7)z CHa. C H [O . CHz .CH(CHa)z]z CH3' CH[O . C H I . CH(CH3) .CHs.CHalz
CH3.CH[O~CHz.CHz.CH(CHa)z]z
C4lIaOz
HzC
CH.HC
I I 0 *CH6CHa). O
I
C&
I
.... .... .'.. .... .... .... ....
64.4 80-90 102.2 146.8 168-170 207-209 210.8
0.821 (22' C.) 0.825 (22" C.) 0.816 (22O C . ) 0.8255 (21° C.) 0.8347 (15' C.)
....
82.6
1.0002 (00 C.)
.,..
85-7
1.118 (16' C.)
94.5-6
....
( i i , 37)
....
O.CH(CH3) .O
....
CsHieOa
201
0.8787( O ' C . )
122-123
0.9641 (0' C.)
224-228
1.027 (0" C.)
(9)
INDUSTRIAL AND ENGINEERING CHEMISTRY
1026
as it does in the case where a monohydric alcohol is involved (reaction 16), it would lead to a splitting off of the polyhydric alcohol-phthalate modifier; the result would then be ureaformaldehyde resin and alkyd resin in a mechanical mixture. Still another type of possible reaction which has not yet been considered is esterification. Since the polyhydric alcohol-phthalate modifier is incompletely reacted (acid number, 150 to 170) a t the time it is introduced into the presence of the urea and formaldehyde, there are necessarily many unreacted carboxylic acid groups. For example, if one molecule of glycerol were absent from the structure proposed in reaction 10, the compound would be a complex polyhydroxy acid instead of a complex polyhydric alcohol. In addition to reacting with the polyhydric alcohol, these carboxylic groups may react with the methylol groups of the urea-formaldehyde complex, thus forming ester linkages. POSSIBLE INTERMEDIATES. The analysis of a typical ureaformaldehyde condensation product consisting of one mole of urea and three moles of formaldehyde in the presence of an excess of butyl alcohol is given in Table 11. These analyses were performed on samples obtained by evaporation of the butyl alcohol a t room temperature over relatively long periods of time. These samples were dried at room temperature instead of a t an elevated temperature in order to obtain an intermediate stage and not to advance the resin to the final irreversible gel state. TABLE 11. ANALYSIS OF TYPICAL UREA-FORMALDEHYDE CONDENSATION PRODUCT Carbon Nitrogen Hydrogen Oxygen (difference)
5 Days 49,90% 18.57 8.43 23.10
10 Days 50.06% 18.46 8.84 23.64
Average 49.98% 18.52 8.63 22.89
In accordance with the analysis of Table 11, the reaction product of urea and formaldehyde in the presence of excess butyl alcohol appears to consist of a partial interaction product of the primarily formed mono- and dimethylolurea and butyl alcohol. Part of the methylol groups may also be converted into methylene groups when reacting the initial urea-
VOL. 30, NO. 9
formaldehyde condensation product with butyl alcohol. The mono- and dimethylolurea might form by interaction with butyl alcohol: NHCHzOCaH9 NHCH20CIHg NHCHzOC4Hg
/
66
co
\NHz
\NHCH,OH b
a 49.3% c 19.2 N 9.6 H 21.9 0
N
47.7% c 15.9 N '3.1 H 27.3 0
N:CH,
)n
C
53.2% 17.7 8.85 20.25
C N H
0
Other compounds present in the resin obtained (before finally being cured-i. e., before heating) might be: XHCHzOH
/
e
d
26.65% 31.1 S.65 3,5.6
C N H 0
33,370c 38.9 N 5.55 H 22.25 O
f 30.0% 23.3 6.7 40.0
C N H
0
Q
h
35.3% c 27.5 N 5.9 H 31.4 0
42.9% C 33.3 N 4.8 H 19.0 0
Formula a or certain mixtures of b and c are very close to the analysis in Table 11. The differences may be explained by slight amounts of e and g, the presence of which is quite probable. On the other hand, it is evident that the intermediate condensation products cannot contain any substantial amounts of d and h. On account of the presence of varying amounts of methyl alcohol in commercial formaldehyde (usually 5 per cent), in the reaction mixture consisting of one mole of urea and three moles of formaldehyde, there would be present one eighth mole of methyl alcohol per mole of formaldehyde which could also react with the urea and formaldehyde in the same manner as the butyl alcohol. Such reactions of methyl alcohol in place of butyl alcohol would tend to give certain discrepancies between the actual analysis and any theoretical formula based on the assumption that butyl alcohol is the only alcohol present which could enter into the reaction.
Manufacture FREE PHTHALIC ANHYDRIDE/GRAM
\
' I
/
FIGURE 1. SOLUBLE, GEL,AND PRECIPITATE AREASOF UREAFORMALDEHYDE MODIFIEDRESINS T h e first number represents t h e percentage of urea-formaldehyde condensation product and the second the percentage of modifying agent; for example, 0/100 = 0 per cent urea-formaldehyde and 100 per cent modifier.
The manufacture (SI) of urea-formaldehyde-polyhydric alcohol-polybasic acid co-condensation products is carried out in two steps-viz., the manufacture of an incompletely reacted alkyd resin modifier and the subsequent co-condensation of urea and formaldehyde with the modifier. Certain combinations of glycerol-phthalate resins in combination with urea-formaldehyde condensation products make admirable light-colored heat-hardening surface-coating materials which are free from the manufacturing and storage difficulties enumerated a t the beginning of this article. Unfortunately not all proportions of glycerol-phthalate with respect to urea-formaldehyde are suitable for the purpose. Quite definite limits exist which spell the difference between success and failure. Between 0-20 per cent glycerol-phthalate and 80-100 per cent urea-formaldehyde, there is a tendency for a white nonresinous precipitate to be formed
SEPTEMBER, 1938
INDUSTRIAL AND ENGINEERING CHEMISTRY
(Figure 1). Between 20-32.5 per cent glycerol-phthalate and 67.5-90 per cent urea-formaldehyde, there is a tendency towards uncontrollable reaction and very rapid gel formation. However, between 32.5-60 per cent glycerol-phthalate and 40-67.5 per cent urea-formaldehyde, a favorable field exists where butyl alcohol solutions of the co-condensation product are compatible with commercial alkyd resin solutions and cheap solvents, such as mineral spirits. In this field the process is controllable in contrast to what occurs in the two areas first described. When the proportions of glycerolphthalate are still further increased between 60-100 per cent glycerol-phthalate and 0-40 per cent urea-formaldehyde, clear pale resins are obtained which are soluble in butyl alcohol but are not compatible with commercial alkyd resins. Some of these proportions may be miscible with commercial alkyd resins in solution form, but in films where the solvents have evaporated they show white because of the incompatibility of the resins themselves. When only a few isolated experiments are made, using various proportions of glycerol-phthalate to modify the ureaformaldehyde, very contradictory results are obtained if the proportions vary over a wide range, and the observer will be apt to think there is no definite relation between the two results obtained. After numerous experiments, however, it is possible to map out the areas shown in Figure 1. The following table shows the results of some of the more pertinent experiments which led to the outlining of the areas in Figure 1: Urea- ModiFormal- fying dehyde Agent
Erpt.
No
% 1 2 3 4 5 6 7
64 100 100 58
90 41 80 40
8
9 10 11 12 13 14 15 0
88
R
n
80
-
70 70 74 40 25 resin:
Acid No."
Acidity Calcd. as Free Phthalic Anhydride
% 46 0 28:OR 0 49R 42 165M 165M 10 59 165M 20 165M 36M 60 100 72M 20 72M 198M 30 126M 30 126M 26 60 165M 75 165M M = modifying agent.
Condition of Resin
Gram
... ...
o:iis 0.218 0.218 0.218 0.048 0.095 0.095 0.264 0.166 0.166 0.218 0.218
OK Clear gel White ppt. OK White ppt. OK Clear gel Clear, straw color Clear no compatibility Wh/tL gel White gel OK White gel OK Clear, no compatibility
The areas shown graphically in Figure 1 are characterized follows:
Range of Free Phthalic Anhydride Type of Area Gram 0/100 to 25/75 0-0.25 Clear, pale resins, not compatible with alkyd resins 25/75 t o 40/60 0-0.25 Clear, sol., compatible with alkyds, decreasing to-no compatibility as 25/75 ratio is approached 40/60to67.6/32.5 0-0.25 Clear, sol compatible with alkyds 67.5/32.5 t o 8 0 . 5 / 1 9 . 5 0-0.25 Clear gei' area with white and cloudy gels od border lines 8 0 . 5 / 1 9 . 5 t o 100/0 0-0.25 White ppt. area: formation of mono- and dimethyl01 compounds Ratio, UreaFormaldehyde Modifier
Within the limits of the favorable area it is possible to make water-white heat-hardening adhesive films which bake out hard and hold their color exceptionally well during baking even if the temperature is as high as 350" to 400" F. The modifier should be prepared in a closed stainless-steel processing kettle equipped with agitator and capable of withstanding vacuum. The glycerol-phthalate compound may be prepared as follows: Three moles (444 parts by weight) phthalic anhydride are heated with four moles (368 parts by weight) of C. P. glycerol at 190° C. until a pill is clear on glass, or until the acid number is between 150 and 170. This usually requires 15 t o 20 minutes at the top temperature for a small
1027
batch. For larger batches a longer time is required. At this stage the modifying compound is allowed t o cool. The resulting thick, water-white, clear, transparent resinous material is now ready for use in combination with the ureaformaldehyde resin. The subsequent co-condensation of urea and formaldehyde in the presence of the modifier just described is accomplished in a processing kettle equipped for indirect heating, agitation, and refluxing, and capable of withstanding vacuum and pressure. RESINA Urea Formaldehyde (37.4%) Glycerol-phthalate interaction product Butyl alcohol
mole 3
60 parts by weight 321 60 222
-
663
This material is refluxed for 2 hours at the boiling point and then vacuum-distilled to a water-free state. The material is adjusted with solvents to the desired nonvolatile content. These resin solutions are usually prepared in 50-60 per cent nonvolatile content.
Properties and Uses of Co-condensation Products There has long been a need for a resin that will give a quick tack-free condition and thorough mar-proofness with a minimum of baking immediately upon removal from the oven together with a corresponding improvement in print resistance. Heretofore, it has been impossible to do this except by the addition of a hard resin, which means an inferior color, but with resin A there is no sacrifice of color. Urea-formaldehyde resins have been on the market for a number of years; while they have been noted for their light color and color retention, they have also had the defects of poor water and moisture resistance, poor adhesion, poor stability, and lack of compatibility and uniformity. In resin A these difficulties have been overcome. This product is a concentrated, heat-hardening, polymerizing resin well adapted to the fortification of softer resins, principally alkyds. It is ideal for baking but is not generally satisfactory in air-drying materials since it is fundamentally a polymerizing resin and not an oxidizing type. It is used chiefly with alkyd resins of the nonoxidizing type and with the medium-oil-length oxidizing types to increase the hardness of baking finishes for metal surfaces. I n combination with resin A the nonoxidizing alkyds give the best color retention, with the oxidizing types ranking second. On account of its pale initial color and lasting color retention, resin A has been found of distinct advantage in the formulation of transparent and colored coatings for metal decorated with light or pastel colors. Its outstanding marproofness also makes it invaluable for the formulation of coatings for metal furniture, toys, motors, and machine parts which have to stand rough handling, especially in transit. Resin A at a 60 per cent concentration in +butyl alcohol is completely miscible with hydrocarbon solvents of both the coal-tar and petroleum types, as well as with stronger solvents, such as ethyl alcohol, butyl alcohol, ethyl acetate, butyl acetate, and the ethers of ethyleneglycol. It is compatible with oxidizing and nonoxidizing alkyds, as well as with nitrocellulose and ethylcellulose. Since there is no discoloration from resin A in the vehicle of an enamel, it is unnecessary to use large amounts of pigment to hide the staining effects of vehicle upon pigment. Consequently, much lower ratios of pigment may be employed, especially where maximum gloss is required. Also, there is little need for using reinforcing pigments because the resin imparts the necessary hardness and mar-proofness.
INDUSTRIAL AND ENGINEERING CHEMISTRY
1028
Metallic driers need not be used with resin A, either when alone or in combination with nondrying alkyds. However, when it is used in conjunction with oxidizing alkyd resine, the usual amounts of driers can be added, if necessary, to meet low-temperature baking schedules.
't
VOL. 30, NO. 9
Coatings containing a relatively high proportion of resin A, when baked a t high temperatures, have good resistance to alcohol, chemicals, and solvents. The inertness of films depends more upon the temperature of the bake than upon the nature of the modifying material. It is desirable to have an inert modifying agent.
The authors are endeavoring to correlate the percentage of alkoxyl groups as determined by the Zeisel method (60) with the nitrogen content to compare with the corresponding values which would be possessed by the hypothetical unit polymer and suggested intermediates. Unit c would have a nitrogen content of 17.7 per cent and an alkoxyl content of 46.5 per cent. The results show an average nitrogen content of 18.4 per cent, but the alkoxyl content is low, probably because of the interference of nitrogen in the compound (60).
./
Summary
PER CENT RESIN A
IN VEHICLE
FIGURE2. EFFECTO F ADDING RESIN A T O A SHORTSEMIOXIDIZING-TYPE ALKYD IN A WHITEPIGMENTED ENAMEL Pigment/aolid binder ratio = 1/0.8
Baking schedules will vary greatly according to the type of product in which resin A is used and also according to the proportions embodied in the finished material. Enamels may be baked on schedules ranging from 1 hour a t 250'kF. (121' C.) to 2-3 hours a t 200' F. (93" C.), but higher temperatures give greater hardness, better waterproofness, and stronger chemical resistance. A bake of 20 minutes to 1 hour a t 300' F. (149' C.) is satisfactory. Baking a t temperatures as high as 350-400" F. (177-204' C.) without darkening is 60-
50
P
I
I
I
I
I
N-CHZL
O
T\I-CHZ--
I
is supported in preference to Redfarn's and Ellis' structural hypothesis on the basis of physical properties of the final resin. 4. Data on structure and physical properties are tabulated for some of the acetals formed as intermediates when the condensation is carried out in the presence of mono- and polyhydric alcohols. For example, n-butyl formal is formed from one mole of formaldehyde and 2 moles of butyl alcohol:
-
40-
30
1. The general reactions of urea and formaldehyde are reviewed and extended to include the acetal and hemiacetal reactions taking place in the presence of monohydric alcohols, and specifically the part played by n-butyl formal in the formation of water-insoluble film-forming urea-formaldehyde condensation products in the presence of butyl alcohol. 2. The reactions of formaldehyde with polyhydric alcohols to form complex ethers and their relations to the cocondensation of urea-formaldehyde resins and polyhydric alcohol-polybasic acid resins are discussed. 3. The reaction mechanism of polymerization resulting in the unit polymer:
-
CHzO
+ 2C4HsOH +CHz(OC4Hp)z + HzO
The general acetal reaction is: SEMlOXlOlZlNG ALKYD X NONDRYINQ ALKYD
Acetal: RCHO
+ 2R10H
----f
RCH(ORJ2
+ HzO
OH
PER CENT
I (20!
' 0 100
80
I
40 60 PER CENT
RESIN A I
60 40 ALKYD
I
I
80
100
20
0
Hemiacetal: RCHO
+ RlOH
/
RCH
\
OR1
RESIN
FIGURE 3. EFFECTOF ADDIXGRESINA TO AN
ALKYD
Baked 30 minutes at 300" F.
obtainable, provided the alkyd employed is a nonoxidizing type. The effect of adding resin A to enamels formulated with a short, semioxidizing-type alkyd is shown in Figure 2. Clear transparent films may be prepared by baking resin A and a short nondrying-type alkyd in various proportions. A similar parallelism is observed when using an oxidizing-type alkyd (Figure 3).
5. The data on a complex polyhydric alcohol made from one mole phthalic anhydride and two moles ethyleneglycol controlled by the acid number and reacted with formaldehyde is discussed. 6. The general reaction of urea with acetals is given ae follows :
NHz d=O
N ' Hz
HNCHRORi
+ 2RCH(OR1)2 +&=O HLCHRORI
+ 2R10H
INDUSTRIAL AND ENGINEERING CHEMISTRY
SEPTEMBER, 1938
7. When the urea-formaldehyde reaction is carried out in the presence of butyl alcohol, on the basis of elemental analysis, the intermediate condensation product is shown to be a mixture of many substances but probably consists largely of: H
NCHxOCdHg I
d=O
I
NCHzOH H
8. The manufacture of urea-formaldehyde-polyhydric alcohol polybasic acid co-condensation products is described. 9. Properties and uses of a urea-formaldehyde-alkyd cocondensation product are given. 10. The effect of increasing percentages of this resin in an alkyd resin is shown graphically, both in clear and pigmented flms. 11. I n clear films the Sward hardness increases from 5 at 10 per cent urea resin-90 per cent alkyd resin to 50 a t 75 per cent urea resin-25 per cent alkyd resin.
Acknowledgment The authors wish to express appreciation to H. Reichhold for his kind permission to publish the article, and also to R. H. Kienle of the Calco Chemical Company, K. P. Monroe and P. J. Ryan of Beck, Koller & Company, Inc., B. W. Nordlander of the General Electric Company, and J. H. Shroyer of Flint Junior College for their kind suggestions and proofreading.
Literature Cited (1) Arnhold, Ann., 240, 199 (1887). (2) Ibid., 240, 200 (1887). (3) Beilstein, Handbuch der organischen Chemie, Vol. I, pp. 575-6. Berlin, Julius Springer, 1918. (4) Ibid., p. 603. (5) Ibid., pp. 603-4. (6) Ibid., p. 604. (7) Ibid., p. 605. ( 8 ) Ibid., Vol. 19, p. 2 (1934). (9) Ibid., p. 11. (IO) Ibid., p. 63. (11) Ibid., p. 64. (12) Ibid., p. 65.
1029
Ibid., p. 436. Ibid., p. 439. Blaise, Compt. rend., 140, 662 (1905). Cheetham and Pearce, P a i n t , Oil Chem. Reo., 99,42,44 (June 10, 1937) ; Oficial Digest Federation P a i n t & V a r n i s h Production Clubs, 8, NO. 167, 216-20 (1936). Chesne, de, Kolloid-Beihefte, 36, 387 (1932). Crump, British Patent 309,849 (1928). DelBpine, Compt. rend., 131, 745 (1900). Ibid., 131, 746 (1900). Dixon, J . Chem. SOC.,113, 238 (1918). Du Font, Compt. rend., 148, 1523 (1909). Einhorn and Hamberger, Ber., 41, 24 (1908). Ellis, “Chemistry of Synthetic Resins,” Vol. I, pp. 576-87, New York, Reinhold Pub. Co., 1935. Ellis, U. S. Patent 2,115,550 (April 26, 1938). Goldschmidt, Ber., 29, 24-38 (1896). Harnitzby and Menschutkin, Ann., 136, 127 (1865). Henry, Compt. rend., 120, 107 (1895). Hill and Walker, U. S. Patent 1,877,130 (Sept. 13, 1932). Holzer, Ber., 17,659 (1884). Hovey and Hodgins, U. S. Patent 2,109,291 (Feb. 22, 1938). John, Ibid., 1,355,834 (Oct. 19, 1920); British Patent 151,016 (1920). Lauter, U. 5. Patent 1,633,337 (June 21, 1927). Ludy, Monatsh., 10, 205 (1889); J. Chem. SOC.,56, 1059 (1889). Luther, Pungs, Griessback, and Heuck, U. 5. Patent 2,019,865 (Nov. 5, 1935). Nef, Ann., 335, 215 (1904). Ibid., 335, 216 (1904). Ramstetter, German Patent 403,645 (1922) ; J. SOC.Chem. Ind., 44, 216T (1925). Redfarn, Brit. Plastics, 5, 238 (1933). Richter, “Organic Chemistry,” 3rd ed., Vol. I, p. 587, Philadelphia, P. Blakiston’s Son & Co., 1934. Ripper, U. S. Patent 1,762,456 (June 10, 1930). Soheiber and Sandin. “Die kunstlichen Harze,” D. 309, Stuttgart. Wissenschaftliche Verlagsgesellschaft, 1929. (43) Scheibler, Trostler, and Schulz, 2. angew. Chem., 41,1305 (1928). (44) Schulz and Tollens, Ann., 289, 27 (1896). (45) Ibid., 289, 29 (1896); Ber., 27, p. 1892-4 (1894). (46) Staudinger, Ber., 59, 3019 (1926). (47) Walter and Lutwak, Kolloid Beihefte, 40, 158 (1934). (48) Walter and Oesterreich, Ibid., 34, 115 (1931). (49) Wurtz, Compt. rend., 53, 378 (1861); Ann., 120, 328 (1845). (50) Zeisel, 3rd Intern. Congr. Applied Chem., 2, 63 (1898); S.Chem. SOC.,81, 318, 115, 193 (1919); see also Kamm, “Qualitative Organic Analysis,” 2nd ed., pp. 206-8, New York, John Wiley & Sons, 1932. 28, 1938. Presented before the Div’sion of Paint and Varnish Chemistry a t the 95th Meeting of the American Chemical Society, Dallas, Texas, April 18 t o 22, 1938.
RECEIVED April
Specific Heats of Organic Vapors PAUL FUGASSI AND CHARLES E. RUDY, JR. Carnegie Institute of Technology, Pittsburgh, Pa.
The Einstein functions used in the specific heat equation of Bennewitz and Rossner are recalculated in the form of the power series, ro FIT rJ2. Use of the r constants leads to a specific heat equation of the conventional form. An example of the use of the I’ constants is given
+
+
I
I
N A COMPREHENSIVE research on the specific heats of orgsnic vapors Bennewitz and Rossner (1) found that the experimental results for a variety of nonlinear molecules containing carbon, hydrogen, and oxygen could be expressed by the equation:
where x q c = No. of valence bonds in molecule i
n = total No. of atoms in molecule Ev,, Es, = Einstein functions for a given bond with characteristic vibration frequencies, v i and 6 i
The numerical values of Y for each bond were obtained from Raman data; the values of 6 were determined empirically from the experimental data.