Isocyanate reactions in and with N,N ... - Wiley Online Library

3 downloads 0 Views 323KB Size Report
1-(3-Methyl)pyrazy1-3-phenylurea. Ib. 11.87 ... Isocyanate reactions in and with N,N-dimethylformamide. (b) at 20 .... Hence, in DMF a significant isocyanate urea.
Die Angewandte Makromolekulare Chemie 191 (1992) 131-139 (Nr. 3418) Zentrum fur Makromolekulare Chemie, Rudower Chaussee 5 , 0 - 1 199 Berlin, Germany

Isocyanate reactions in and with N,N-dimethylformamide Detlef Joel, Petra Miiller, Rosemarie Ah1 (Received 9 August 1991)

SUMMARY Solutionsof phenylisocyanate (I) in DMF were stored at different temperatures in the presence of both polyurethane catalysts and 1,3-diphenylurea (11). The reaction products were determined by HPLC. Beside the hydrolysis of I to 11, a significant reaction of I with I1 to 1,3,5-triphenylbiuret(111) was observed at 20°C. Above 60°C I reacted with the solvent to form N,N-dimethyl-N'-phenylformamidine(V). In DMF, 111 dissociated to I and I1 already at temperatures >40"C. ZUSAMMENFASSUNG: Liisungen von Phenylisocyanat(I) in DMF wurden sowohl bei verschiedenen Temperaturen als auch in Gegenwart von Polyurethankatalysatoren und/oder 1,3-Diphenylharnstoff (11) gelagert. Die Reaktionsproduktewurden mit Hilfe der HPLC quantifiziert. Neben der Hydrolyse von I zu I1 wird bei 20°C eine signifikante Reaktion von I mit I1 zu 1,3,STriphenylbiuret(111) beobachtet. Oberhalb 60 "C reagiert I mit dem Liisungsmittel zu N,N-Dimethyl-N'-phenylformamidin (V). In DMF dissoziiert 111 bereits bei Temperaturen >40 "C zu I und 11.

Introduction N,N-Dimethylformamide (DMF) has been applied in the polyurethane chemistry for a long time, especially in the synthesis of polyurethane ureas. Because of its importance for the production of polymers, special purification methods have been developed for this polar solvent'. Unexpectedly, after a prolonged period of storage, NCO/DMF solutions showed essentially higher NCO conversions than expected by traces of water2. To clarify details of the considered NCO reactions, we investigated solutions of phenylisocyanate in dependence on time at 20°C, in the presence of catalysts, in the presence of 1,3-diphenylurea (11) as well as in dependence on the temperature. After blocking the free NCO groups with methanol yielding 0 1992 Huthig & Wepf Verlag, Base1

CCC OOO3-3146/92/$05.00

131

D. Joel, P. Muller, R. Ah1 N-phenylmethylurethane (Ia) or with 3-(5)-methylpyrazole yielding I-(3methy1)pyrazyl-3- phenylurea (Ib), the possible reaction products listed in Scheme 1 and Tab. 1 were quantitatively determined by means of HPLC3s4,5. Tab. 1.

Retention time and calibration constants of the reference compounds separated on a RP- 18 column (5 pm).

Compound N-Phenylmethylurethane 1-(3-Methyl)pyrazy1-3-phenylurea 1,3-Diphenylurea 1,3,5-Triphenylbiuret Triphenylisocyanurate 1,l -Dimethyl-3-phenylurea 1-n-Butyl-3,5-diphenylbiuret 1-n-Butyl-3-phenylurea Naphthalene N,N-Dimethyl-N'-phenylformarnidine RP(0DA)A; 10 pm

Ia Ib 11 111 IV VI VII VIII IX V

t, (min)

Amount/area

3.27 11.87 7.55 23.49 14.62 1.87 23.48 4.93 21.14

5.5933.10-7 2.92485. 1.7381 . 3.1691 . 1.4427 . 10 - 5 9.4815 * 5.6417. 1.1297 * 1.3152.

5.8

2.7992 *

Experimental Reagents N,N-Dimethylformamide was distilled in the dark through a 60 cm Vigreux column at a pressure of 16 Pa under nitrogen atmosphere. Then it was stored for 14 days over a molecular sieve 4 A. The specific conductance of the solvent was less than 8 x lo-* L 2 - l crn-I. The water content determined at the beginning of the experiments was 0.032 wt.-To. During storage, however, it increased to 0.05 wt.-To. Dimethylamine could not be detected by gas chromatography. Phenylisocyanate (I) (p. a., Merck) and the catalysts 1,4-diazabicyclo[2,2,2]octane(DABCO), dibutyltindilaurate (DBTL) and bis(tributy1tin) oxide (BTO) were used without further purification. The model compounds in Tab. 1 were synthesized and purified in the known way. N,N-DimethylN'-phenylformamidine (V) was prepared according to6*7 . The purity of the compounds included in Tab. 1 was checked by HPLC.

Procedure and preparation of samples 0.5 molar solutions of I in DMF were stored under the following conditions: (a) at 20°C,

132

Isocyanate reactions in and with N,N-dimethylformamide

(b) at 20 "C in the presence of 0.05 mol-Vo DABCO or DBTL, (c) like (b) with addition of 0.1 mol/l of 11, (d) like (a), depending on the temperature. After defined storage periods, samples of 1 ml were taken. For blocking the free NCO groups they were treated with methanol yielding Ia or with 3-(5)-methylpyrazole yielding Ib. Scheme 1. Isocyanate reactions in and with DMF.

I

+ R-NH,

R-NH-CO-NH-R (11)

I

+ I1

R-NH-CO-N-CO-NH-R

I

R (111)

-

3 I

0

II

R-N

I

/c h - R

(3)

I

o=c, ,c=o N

I

0

I

+ (CH,),N-C II

A

I

R\

N=C=O

.1t

H

(CH,),-N-C=O

-co,

b

R-N=CH-N(CH,), (V)

(4)

I

H I

+ HN(CH,),

__*

R-NH-CO-N(CH,), (VI)

133

D. Joel, P. Miiller, R. Ah1

Chromatographic conditions A Hewlett-Packard H P 1084 A liquid chromatograph equipped with a 254 nm fixed wavelength UV detector was used for analysis. With the exception of V, all compounds included in Tab. 1 and Fig. 1 were separated on a columns (100 x 4.6 mm I. D.) packed with endcapped 5 pm RP-18 silica 100. The eluent was a mixture of double-distilled water (A) and acetonitrile (B) for UV spectroscopy (PCK Schwedt).

HI

m

Fig. 1. HPLC chromatogram of the model compounds; VI: 1,I -dimethyl-3phenylurea, Ia: N-phenylmethylurethane, 11: 1,3-diphenylurea, IV: triphenylisocyanurate, IX: naphthalene, 111: 1,3,5-triphenylbiuret. 8

16

t,(rnin)

2L

The chromatographic separations were achieved by means of a gradient program: First, isocratic elution was carried out up to 16.5 min at 36 vo1.-Vo B, then a gradient elution followed from 36 to 60 v01.-Vo B up to 20 min and, finally, an isocratic elution at 60 vo1.-To B up to 24.5 min. After elution, the column was washed with acetonitrile for 3 min. The flow rate was 1.2 ml/min. The reactivity of V causes interactions with the residual silanol groups of the RP-18 silica and the resulting tailing peak cannot be quantified exactly. For the quantitative determination of V, a 140 x 4 mm I.D. column packed with RP(0DA)A sorbent was applied. RP(0DA)A is a silica gel (LiChrosorb Si 1 0 0 , l O pm) modified with polybutadiene epoxide and n-octadecylamine. A constant mixture of 60 v01.-Vo water and 40 v01.-Vo acetonitrile at a flow rate of 1.O ml/min was used as eluent. The separation temperature was 40°C. 10 1.11 of the sample solution was injected each. Internal standard was naphthalene. The chromatograms were evaluated quantitatively with a Hewlett-Packard 79850 A LC integrator. The quantification of V was achieved by determining the difference of the sum of reaction products to 100% and by direct quantitative analysis on a RP(0DA)A column with water and acetonitrile as eluents. The results of the two methods were in good agreement. However, after several analyses the retention time of V decreased. This effect will be investigated in future.

134

Isocyanate reactions in and with N,N-dimethylformamide Results and discussion

NCO conversion at 20 "C The reaction products formed at 20°C in DMF are compiled in Tab. 2-4. According to these results, during the first 24 h the decrease of NCO can be explained only by the reaction with water leading to the formation of I1 according to Eq. (1). After this time, a small amount of I11 was registered additionally. The concentration of I11 rises with increasing time. Within 384 h, Tab. 2. NCO conversion in DMF at 20 "C in dependence on time. Time (h)

Ib

I1

Conversion (070) 111

IV

2 24 48 120 384

95.7 90.4 88.8 84.4 53.1

5.3 8.2 8.6 6.7 4.2

-

-

-

-

-

-

1.8 6.9 11.0

-

-

1.5 15.6

1.3 5.8

VI

Tab. 3. NCO conversion in DMF at 20 "C in the presence of 0.05 mol-Vo catalyst. Time (h)

Catalyst

24 96

DABCO DABCO

24 48 96

DBTL DBTL DBTL

Ib

I1

85.3 73.9

8.2 2.1

73.4 27.2 12.5

3.6 3.3 2.5

Conversion (070) 111 IV

VI

-

-

3.3 14.8 12.3 12.8 13.8

-

1.7

3.8 49.1 61.7

-

1.9

11'-70of I is converted into 111. Hence, in DMF a significant isocyanate urea reaction should be considered even at 20 "C. In contrast to Jovtscheff*, in our experiments the cyclotrimerization reaction occurs relatively late. After 120 h only 1.5% of I and after 384 h 15.6% of I are converted into IV. According to Ulrich9, DMF itself may be a catalyst for the trimerization of aryl isocyanates after prolonged storage at room temperature.

135

D. Joel, P. Miiller, R. Ah1 Tab. 4.

NCO conversion in DMF at 20 "C in the presence of 0.1 mol/l of I1 and 0.05 mol-Vo DBTL or BTO.

Time (h)

Additive

Conversion (070)

Ia

I1

111

IV

VI -

~~

24 48 72

I1

24 48 72

I1

+ DBTL + BTO

32.2 14.3 4.0

25.7 32.1 37.0

33.6 46.5 54.8

0.3 0.3

9.4 5.2 4.5

23.5 25.6 26.2

60.3 62.2 64.7

0.3 0.3

-

In our experiments, however, IV is formed only after some days, when further NCO derivatives (11, 111) are formed. We can, therefore, assume other species to start the trimerization reaction. I,l-Dimethyl-3-phenylurea(VI) can also be detected only after prolonged storage. Up to 120 h a reaction of I with the solvent could not be observed. Therefore, we explain the formation of VI by partial degradation of the solvent. During storage at 20 "C for 16 days, no uretodione and no carbodiimide have been observed. Recently, similar results have been reported by Suzuki et a1.I0, who studied side reactions of I in N,N-dimethylacetamide (DMAC). To minimize the undesired reactions in polar solvents, such as DMF and DMAC, the synthesis of polyurethanes should be carried out at low temperatures and within short time. In the presence of the well-known polyurethane catalysts DABCO and DBTL, the NCO concentration decreased more rapidly (Tab. 3 and 4). In these cases, substantial amounts of I11 were formed already after 24 h and, in the presence of DBTL, IV was detected additionally. In the presence of DBTL, the conversion of NCO was higher than in the presence of DABCO. The activity of the two tin catalysts DBTL and BTO is apparent particularly at a higher urea content. As shown in Tab. 4, under such conditions already after 24 h the total amount of I1 is converted into 111 and the greater part of I reacted to IV. For the time being, it is not possible to deduce the differences in the effect and the selectivity of the two catalysts DBTL and BTO from these results. The solution of this question requires further experiments varying the reaction time, the concentration of reactants and the catalysts.

136

Isocyanate reactions in and with N,N-dimethylformamide

NCO conversion in dependence on the temperature To examine the influence of temperature, the solutions of phenylisocyanate were heated at temperatures ranging from 40 to 120"C at intervals of 20 "C for two hours each. Samples were taken at every interval, one for HPLC analysis and the other one for determining the NCO content by means of din-butylamine titration. The results are illustrated in Fig. 2 and Tab. 5. Beside the expected 11, further reaction products did not occur up to 40"C.The NCO conversion increased rapidly with increasing temperature.

Fig. 2. NCO conversion in dependence on the temperature; reaction time: 2 h. For explanation of symbols see Tab. 1.

Tab. 5. NCO conversion in DMF after 2 h in dependence on the temperature. Temperature Ia ("C) 20 40 60 80 100 120

95.1 90.6 7.3 62.6 51.3 37.3

Conversion (Yo) I1

I11

IV

V

VI

6.7 8.8

-

-

-

-

-

-

4.4

11.2 9.1 7.2 2.8

3.1 12.2 13.7 13.3

1.85 7.2 15.1 35.3

0.18 0.95 0.9 0.9

5.3 6.3 9.8

-

137

D. Joel, P. Miiller, R. Ah1

At 6 0 ° C 11% of the initial concentration of I was converted to 111. However, above 60 "C the concentration of I11 decreased again and the concentration of I1 increased correspondingly. This indicates that the "thermal" dissociation of 111 in DMF is favoured. Further NCO conversions above 60 "C can be explained by trimerization to IV and, to our surprise, they can also be ascribed to the reaction of I with the solvent affording N,N-dimethy1-N'phenylformamidine (V). Whereas 1.8% of I reacted with the solvent at 60"C, as much as 35% of I was converted this way at 1-20"C. Losses of I due to the formation of VI are relatively small and the amount is less than 1% even at 120°C. The favoured dissociation of I11 in DMF was confirmed by heating solutions of 111 and, for comparison, the corresponding solutions of 1-n-butyl-3,5-diphenylbiuret (VLI) for one hour at 40, 50, 70, and 90°C. Subsequently, the solutions were rapidly cooled below IO"C, methanol was added, and after a corresponding dilution they were chromatographed. The results of this series of tests are presented in Fig. 3. They confirm that in the case of 111 a reverse reaction can be expected already at temperatures above 40°C. At 9 0 ° C the equilibrium according to Eq. (2) lies largely on the left side. In contrast to that, the partially alkyl-substituted biuret (VII) is far more stable. The dissociation into I and I-n-butyl-3-phenylurea (VIII) was found to start only above 90°C.

Fig. 3.

138

Dissociation of I11 in DMF to I and I1 in dependence on the temperature and for comparison with VII; reaction time: 1 h.

Isocyanate reactions in and with N,N-dimethylformamide Tab. 6. Dissociation of 111 in DMF depending on the temperature. Temperature ("C)

I11

I1

40 50 70 90

94.5 88.9 73.3 8.2

3.6 7.4 19.6 62.3

a

Conversion (070) Ia 1.8 3.3 7.9 30.6

(I)a (1.9) (3.4) (8.9) (30.3)

Titration of NCO.

The authors wish to express their thanks to Mrs. J. Reich for her technical assistance in HPLC measurements and to Mr. U. Erler, Institute of Organic and Macromolecular Chemistry of the Friedrich Schiller University Jena for leaving the RP(0DA)A material.

' lo

J. Juillard, Pure Appl. Chem. 49 (1977) 885 A. Jovtscheff, F. Falk, J. Prakt. Chem. 13 (1961) 265 D. Joel, H. Much, Plaste Kautsch. 32 (1985) 11 S. W. Wong, K. C. Frisch, J. Polym. Sci., Part A: Polym. Chem. 24 (1986) 2867 D. Joel, W. Hettrich, P. Dietrich, Fresenius Z. Anal. Chem. 330 (1988) 136 M. L. Weinert, J. Org. Chem. 25 (1960) 2245 H. Bredereck, F. Effenberger, G. Simchen, Chem. Ber. 98 (1965) 1078 P. Dietrich, A. Kunath, B. Hoffmann, E. ARmann, Z. Chem. 28 (1988) 428 H. Ulrich, J. Polym. Sci., Macromol. Rev. 11 (1976) 93 T. Matsui, H. Karnatani, Y. Arimatsu, A. Kaji, K. Hattori, H. Suzuki, J. Appl. Polym. Sci. 42 (1991) 2443

139