T A B L E \' ABSULGTF, \.MRDET c O S s l . . l S T S O F 13ESZfi.VL '?unp 1.
A
2
,XI nl@
"C Density
1. X 10' T'/d X 1 0 2
cj0,89474
3.0412 3.0175 2,9940 3.9705 2.9470 2.9233 2.8763 2.8296 2,7826
10 13
3! 2,; 30 40
50 ii0
,88935 ,88397 .8785!) ,87321 ,86783 .85706 ,84629 ,83552
X V/d,.U
2
X 10'
3.3991) 2.6550 3.1759 : 3 . 5495 3.3929 2.6503 3.1512 :i.5133 3,3870 2.6456 3,126(i : i .5371) 3.381C) 2.6409 :3.1090 3.330ti 3,3749 2.6361 3.077:3 :i.5241 3.3688 2.6314 : ~ . 0 3 7 :3.5177 3.3569 2.6213 3.00;15 3.2044 3.3435 2.6116 2.9542 3.4908 3.:3304 2,6011 2.9049 3 ,4767
less than those for any previous \;erdet constants. I t is seen from the data in Table IV that the dispersion for benzene is not constant but decreases slightly with temperature. This behavior is similar to that found previously for the lTlagIletic rotatory dispersion of water above 25'.18 The in Table v show that the molecular Verdet constants, A l ~ b ~ / cdecrease j, with temperature o\.cr the eIltirc range. I ~ i c l l a r d s o n ~ ~ has (13) Chas. E.\\:aring and R r h e r t 1,. C u s t c r . '1 H I S . l o t i ~ x . 74, ~ ~ 2306 , ,' 1'3.3,).
DEPARTMEST OF
ISDUSI'RI.11,
V
3 . 7725 2.i677 2.7627 9 .7577 2.7527 2.7477 2 , it372 2.7267 2.7156
A -. l . $ l i m ~
V X
.M
10' L"iI
X 10:
3.6295 .4.0563 3,1085 0.3568 7 , 104G :S.W19 4.0494 3.1830 6..3069 7.0910 3,5731 4 0431 3,157:j 6,2571 7 . ! ) i 8 4 :3.5449 4,0848 3,151ti 6,2073 7 .(1651 3.5167 4.09i;Y 3,1457 6,1574 7,!1314 3.48% 4.0199 3.1399 6,1076 7 OSTH 3.4321 4.11045 3 . 1279 t j ,0079 7.0099 3.3757 3.9888 3.1156 5,9083 6.9814 3.31'33 3.9727 3,ll):i1 5,8086 6,9381
1
3.549-4 5.5392 5.5289 5 51% 5,507S j.4978 5.4754 3, 5,430t3
suggested that the origin of the temperature dependence of the normal Faraday effect might be in the thermal opposition to the precessioml motion Of finite magnetic nloments, thus, to a mean paramagnetic j'olarization. These results, therefore, might be interpreted in terms of a slight p a r a magnetic polarization in the case of benzene. I t is of interest to note that the temperature coefficient a t 430 nip is approxiniately twice that at 589 1 1 1 ~ . Further extensions of theory would have to account for this behavior. SI IJKR5,
( 1 * l i b. S. Richardson. PI-OLRoy. SOL. Loudoizj, 31, 232 ,d!!Ilii
[ C O S T R I B U T I O N FROM THE
x =x i flLL, 1' x IO' V / d x 102 I
378 r n w
1. X 10: 1"d
C O S S E C I I C [ . I'
C H E M I S T R Y , THE FACULTY OF E N G I N E E K I S G , K Y O T O r S I V E R b I T Y ,
Kinetics of the Condensation of Melamine with Formaldehyde BY ~ I A s . \ YO~.\n-o ~\ A N D YOSIIIRO OG.~T,\ RECEIVED DECEMBER
26, l!Jjl
The rates of the condensation of melaminc with formaldehyde in aqueous media in thc PH's varying from 3 to 10.6 have been estimated a t 35, 40 and 70' by employing both iodometric a n d sulfite methods. The primary product of the condcnsation, methylolmelamine, was found t o consume iodine like formaldehyde itself, and therefore the result of the iodometric method indicates the amount of formaldehyde together with that of methylol group present, irhile the sulfite method estimates solely formaldehyde unrcacted. From the data obtainctl by thcie two methods, the rates of the fxmation of methylolmelamine, e.g., ( I ) , and methylene-bonded melamines, c . ~ . (, I 1 J , r r e r e determined. There were found evidences that only the formation of methylol conipounds occurs at 35-40" except i n ;iridic solutioii and that the reaction is reversible all through the p H range, its forward rate being proportioual to either [ m e l m ~ ~ n[formaldehyde], c /melamine] [conjugate acid of formaldehyde], or [conjugate base of melamine] [formaldehyde] accordiiig to t h e PH's of the media. The rate of the irteversihlc condensation of methylomelamine with melamine in ricutral and acidic media at 7(Jo, where the hydroxymethylation ( I I i . very rapid, is determined by the condensation step of conjugate acids of li~ethyloltnelamineswith melamine, the rate being expressed as [melamine]* [formaldehyde]. The rate equations based 011the reaction mechanisms which are suggested froiii these results satisfied the relationship bet1veen p€i and the rate co~lstantexperimentally fouud. ,
It has been pointed out that the nature and structure of the melamine-formaldehyde condensate are considerably varied with reactioii conditions, especially molar ratio of reactants, pH and temperature of the solution.' There seems to kit. no
m c s
s
decisive kinetic evidence for this condensation mechanism, although the primary process inay be presumed from the similar reactions such as those Of urea-formaldehyde and aniline-for~naldehycle.' ( 2 ) T S H o d g i n s a n d h I' H o \ r \ l l c i u i ~ r r r i .1rln 1 8 , 431 ( 1 % ? J
ihi
/
3 0 , 10.21 I ' l 3 R
K I
ri I
wov. 20, 193%
KINETICSOF
THE
CONDENSATION OF MELAMINE WITII FOKRIALDEIIYDE 5720
tion, the resinous product was filtered and dried under reduced pressure. The product forms white amorphous powder, insoluble in water, alcohol and benzene but soluble in strong acids and phenol, just as the aniline-formaldehyde condensate, and softened at about 300' with decomposition. This substance was dissolved in aqueous acid and tested with two analytical methods, of which the sulfite method was utterly insensible t o it, but the iodometry revealed the a presence of some unattacked methylol groups in the moleExperimental cule Experimental Results and Calculations.-The rate of Materials.-Xippon Carbide Ind. Co. best grade melamine was recrystallized from water, m.p. 353" (cor., hydroxymethylation (I) may be calculated from the differblock). The formaldehyde was the best grade. I t was ence of two values obtained by sulfite and iodometric methalso confirmed that impurities in the formaldehyde used, ods; however, i t was recognized that in the pH range above 6 or below 4 no methylene bridge formation but the hydroxye.g., formic acid or methanol, did not affect the estimation methylation was possible a t the temperature ( 3 5 and 40'). of the reaction rate a t least in their small quantities. A Typical Procedure for the Rate Measurements.-In an The hydroxymethylation is reversible and resembles that erlenmeyer flask or a flask equipped with a reflux condenser of urea.' Since trimethylolmeIamine may easily be obfor the higher temperature experiments, 2.5226 g. (0.02 tained as a main product, it appears probable that there is hardly difference in the rates between the hydroxymethylamole) of the melamine was dissolved in distilled water, added with appropriate amount of dilute hydrochloric acid or so- tions of melamine, mono- and dimethylolmelamines. Consequently, the rate of the consumption of formaldehyde v~ dium hydroxide, diluted to 180 cc., and dipped in a thermostat regulated to the operating temperature, Into this may be expressed as solution was poured 20 cc. of 1.00 A/l aqueous formaldehyde ~ I A= k(3m - x ) ( f - X ) - k ' x (1) solution, maintaining the same temperature, by using a caliHere, k and k' are the forward and reverse rate constants, brated pipet. A twenty-cc. aliquot of the mixed solution respectively, nt and f the initial concentrations (moles/l.) was pipetted out a t known intervals of time and estimated by the sulfite method and a 5-cc. aliquot by i ~ d o m e t r y . ~of melamine and formaldehyde, respectively, and x the conI n the sulfite method, the aliquot was run into 25 cc. of 0.5 sumed concentration of formaldehyde after t seconds. The N sodium sulfite solution and the produced sodium hydrox- integration of the equation leads to ide was immediately titrated with 0.1 A; hydrochloric acid by using rosolic acid as an indicator. When a n acidic or alkaline reaction medium was employed, it was necessary to add previously to the sulfite solution an appropriate amount of acid or alkali for the neutralization. In the iodometry, the aliquot was poured into ca. 10 cc. of 0.5 N sodium hydroxide, added 25 cc. of 0.05 N iodine solution, where xB is the n a t equilibrium state, which may be determined graphically by plotting the conversion percentage v s . then i-8 cc. of 1 N hydrochloric acid after five minutes, time. Table I ( a ) lists the rate constants k and k' with and thereafter titrated with 0.05 Ai sodium thiosulfate. The Reaction Products Criterion and Blank Tests. TABLE I ( A ) Trimethylo1melamine.-Since this substance was solA N D THE uble and could not be isolated from such a dilute aqueous THERATE CONSTANTS O F HYDROXYMETHYLATION solution as that used for rate measurement, it was synthe- FORMATION OF METHYLESEBRIDGE: THEEFFECT OF THE sized by the usual method,4in which 1 mole of melamine was MOLARRATIOO F REACTANTS stirred with 3 moles of 35% formaldehyde of PH 8.0 a t 70' ( a ) Hydroxymethylation until the dissolution was complete, heated for additional ---Initial concn.-k'av five minutes, cooled readily, filtered, washed with meth.M .M Formal!fa,x 10' x anol and then dried in vacuo. This crude product, conTemp., Melamine dehyde (I.*mole.-' 104 "C. (in) (f) pH sec.-I)a (sec.-l) taminated with a small amount of melamine, higher condensate and formaldehyde was difficult to purify; it was 40.0 i 0 . 1 0.100 0.0;10-0.300 7 . 7 2 , l 4 & 0 03 0 . 4 3 dissolved in warm water, filtered and evaluated with both 40.0j; , 1 0.050-0.125 0.100 7 . 7 2.113- ,04 46 sulfite and iodometric analyses. The results indicated 4 0 . 0 i .1 ,030- ,150 .IO0 3 . 8 0 . 7 2 zt .U1 .76 that the sulfite method measures solely the formaldehyde 35.0 r , I ,035- ,063 ,100 1 0 . 2 6 . 2 32 , I .56 present, whereas the iodometry cap estimate not only for( b ) T h e formation of methylene bridge maldehyde but methylol group present mainly as trimethylkoa\, X 103 olmelamine5 in this case. I t was also confirmed that even (I.l.mole -2,sec. -1) a a larger excess of formaldehyde, e.g., 2.0 M formaldehyde 70.0 f 0 . 2 0.100 0.050-0.300 4.9 2.67 h 0.04 per 0.1 ,If melamine, does not increase the number of methylol group per one melamine molecule more than three, a t Figures following f mean probable errors. least under the conditions similar to those of kinetic experiment ( a t 40 and 70" in neutral media); therefore, the most TABLE I1 hydroxymethylated state is surely trimethylolmelamine.~ THE RATE CONSTANTS O F HYDROXYMETHYLATION O F (B) Methylene-bonded Melamine.-Even a dilute soluMELAMINE:THEEFFECTOF PH tion as used in the rate measurement emulsified after cn. 20 (a) Alkaline media a t temp. (b) Neutral a n d acidic media rninutcs a t i o 0 and pH about 5. After allowing to stand 36.0 i (1.1' a t temp. 40 0 i 0 . l o over threc hours a t the temperature to complete precipitaInitial concn.: melamine, 0.0: .bf; Initial cuncn.: melamine, 0 I O Jf:
The present report summarizes our results on the melamine-f ormaldehyde conden sation studied from the kinetic standpoint, placing emphasis upon the important relation between pH and the rate, which implies some probable mechanisms for the early stages of the condensation.
(I
(8) J . F. Walker, "Pormaldehyde." Reinhold Publishing Corp., New York, N. Y.. 1044, pp. 266. (4) E . g . , (a) J . K. Dixon. N. T. Woodberry a n d G. W. Costa. THIS J O U R N A L , 69, 500 (10471, (b) C. Kitagawa, J . Chein. SOL.J a p a n , 63, 181 (1950). (6) Hexamethylolmelamine is also known. However, i t was impossible t o isolate this material beside trimethylolmelamine under these conditions. I t would be formed only a t higher temperature a n d concentration. (6) It was observed t h a t in the sulfite method standing for a longer time or estimation of aqueous methylolmelamines in alkaline medium remarkably increased the titer of hydrochloric acid, while in the iodometric method the titer held constancy. These phenomena seem t o indicate t h a t alkali. which i 5 either originally present or produced by the reaction of sulfite. accelerates the decompositirin of methplolmelamines into mslamine and formaldehyde ( c f . equation 0 ) .
PH
formaldehyde, 0.10 .M k X 104 (l.~mole-J*sec.-~)a
formaldehyde, 0.10 .M k x 10' il:mole-'.sec.
PH
7.6 1.62 f 0 . 0 1 7.7 2 . 1 4 f 0.08 9.U 1 . 8 9 f .02 5 8 2 . 1 3 f .02 9.5 2.54 f . 0 3 5.3 1 . 9 4 i .03 10.0 3 . 7 6 f .04 4.9 1 . 3 5 i .02 10.2 6.3 f .1 3.8 0 . 7 2 f .01 10.4 8.9 f .2 3.0 0 . 7 4 f .01 10.Fj 11.7 f .2 10.6 14.1 f . 2 Figures following f mean probable errors. (71 G. A . Crowe, J r . , a n d C. C. Lynch. T H I S J n u R N A i . . 70, 371)s (1048).
varying molar ratio of reactants. Table I1 shows the of the conjugate base of melamine with formaldeeffect of pH on the value of k with a constant molar ratio. On thc other hand, t h e rate of methylene bridge formation hyde, that of melamine with formaldehyde and that of melamine with the conjugate acid of form(11) may be directly evaluated iodometrically. I n order t o aldehyde (IS. equations 6, 1.5 and 12). simplify the analysis of the rate equation, the rate was measured at 70", where the hyclroxymetl~ylation was so Like the urea-formaldehyde condensation, l o the rapid t h a t the reaction mixture soon reachecl to a mobile hydroxymethylation in alkaline media appears to cquilibrium. Since a reaction between aminomcthylol and amino group is rate-determining, the rate may be given, irwoh-e an attack of nitrogen anion of melamine 011 the carbonyl cdrbon of formaldehyde. This mech:IS t h a t iii tlic aiiiline--forinaldeliy~~e contlciisatioti,~in a form anism will lead to the rate equation previously :'I3 = k " ( 3 ~ ~z i -- zy,x(,f -- i - y ) Here, k" is the rate constant, y the concentratio:i (molesil.) described. 'The iiiechaiiism involves the stages of formaldehyde consunicd for the niethylene bridge forimSI11 011-11IilO (fast) (5) tion after t seconds, z the conceiitration of fortnaltlchytlc S I 7IICHO MCI-IAO(slot\) (6) coiisumed for t h e hydrox?micthylation. As tlc.;cribeti above, z may be dcterminetl from the difference in t w o ; ~ m ,11Cll~O -/- ILL) XIC€I,OH -1 O H - (f,tst) (7) lytical methods; this v;tluc 1)ccaine it. coiistant \eon aftcr where hII-I is a ~iiolcculeof iiielaiiiine, M- tu anion starting the reaction arid teiitls to dccrc;tsc slightly wit11 the proccssioti of the contictisation, lint t h e tlccrcasc was ticg- of riielaiiiiiic. Equilibriuin (5) is probable because ligihly small during the rate ttie~tsiiretiic~tl.~' The r:ttc coiiiiie1,iminc is ablc to form iiietallic saltsi and furtherstant k" iii:ty bc calcu1:ttctl liy intcgratiiig equntioii ( S ) :I\ I
+-
-
fOllO!\S
'Tlicsc k" values with v:trious iiiolar ratios at it coiistant pI1 are shown in Table I ( b ) , which proves the adequacy of equation (3). Although it ~vould surcly bc desirable to list also t h e data with varying concentrations of rnclarnirie, the change of the concentration of basic inrlairiinc in weakly acidic solution brought about :i cotisider:rhle change in jOH, which also influeiiccd the reaction rate to it litrgcr extent ;is described below, arid hence this experiment was :thantlotictl. The values of k" tabulated werc calculated frotii the data with less than 35% conversion of formaldehyrlc. They hold fairly good constancy within this rangc., but tciiti to rlccrease as tlic condcrisation proceeds over :mil it become.; apparently more accurate to replace the tcrm ( 3 ~ n- i 2 y ) in equation ( 3 ) by (3vz - z - 5 ~ ) .I t thus appears that there is very small range in which ( 3 -~ :-- 3 y ) or ( 3 m z - 4y) may- be applied; thcse facts would be accoutitetl for b y assuming t h a t the three ainiuo groups in :t inclnmiiic molecule have thc snnie reactivity for rncth~~lol group :IL early stages of tlic conderisatio~~ i n riilrite soltition, lxit : t i the reactioti proccctls the rcactiv falls r:tI)i(Ily by tli(> production of stcrically Iiiiitlcrctl a i 1 ( 1 g r o i t p i i i the C O I I tlcnsatc. The effect of pl-i on the rate \vas roughly irie:isurerl by coniparing thc reaction ratio after definite timc, because the solution soot1 bccatne turbid tir the re:rcti slow. Tile pIi values listed in these tuhle at 25' b y incalls of a glass clcctrotle l.il3I.E
THE COMPARISON OF
111
RATES01: T I I B PORMATIOS (it: ' r l % EL':FFIIC'I. O F />€I A 7 rrEJI€'. 70 k 1 TIIE
LfETEiTLENE RRIDGF;: Initial concn.: nirlqniinc, 0.10 .If; furmalrlchyrlc, 0.10 .!l' Conversion, Yea Convrrsion, < hftw Af t r r After Alter pi1 1 . j inin :iO m i t i PI1 2 0 milt. (XI r n i i i .
!;. 21; 1 .:< 1; ..ill :; , x "I I 4.9 11 2( I These values S h O W the allloUtlt Uf
5 .s 5,5 5.2
'I
I!)
1:: )
3
.I-
.)
C l J l l S L l l l l ~ d forlll:l~dC-
hytlc which is riieasured iodometric:tlly, i . c . , thc coitvcrsioii percentage of formuidehyclc into -CHn- bridge.
Discussion of Results There seem to be three probable mechanisms as to the formation of methylolmelaniines correspondextion media, ;.e., the reaction jlrsti
.\I 5 1 1 g x \ \ . ~ r '1 a ,in:;J O I I R ~ A I . , 73, 171
-)
T h i s would lie a t t t i b u t e r i t o t h e fact thxt ainifio jirriups i n t h e rondcnsrtc of loivcr molrc \ % r i g h t i r e A - ?,i>iI\. 1ivtlro~virir.Iiiyl :itcii
.is tIiuar
i t 1 iiicI.iiii!iii
I
more tlic sulutioii o f i t in dilute aqueous alkali shows pH which is consistent with this cquilibriuiii (K,l = LLL. 10-l2)). I t has been known from polarographic iiivesti~atioiisi",llthat the degrec of hydration of forinaldehyde, is.,the cquilibriuiii constaiit of equation HCHO I320 P? CH2(0HJ2,is iiiuch influenced by $13, especially in the alkaline medium. But there seem to be no available data on thc hydration degrec in such a high ccxicentration of aqueous formaldehyde as those in our experiiiients and also our attempt to obt:iin these data was unsuccessful. The following rate equatioii riiay thus be obtained by assuming that there is no shift in the equilibrium above, the equation lioldiiig a t least in the relatively narrow pII r'tiige.
+
11 = -
k b ; W
ks(:j/rz
I:cII?o]- t-s[,lICHO-] - N I ( _~ - X_ ) -~ k- 62: I _
[Ii:OI IK,[OII-I + k ( i i i i - Z,(J.
-
' \ -- k'xi ' -'+ \
/
\
~
K,[HzOI/ [OH-] \
S)
(8)
wlim I
