1888 R. J. SLOCOMBE, E. HARDY,J. H .

SAUNDERS
EDGAR AND R. L. JENKINS Vol. 72

[CONTRIBUTION
FROM THE RESEARCH PHOSPHATE
LABORATORIES, DIVISION,MONSANT?
CHEMICAL
COMPANY
]

Phosgene Derivatives. The Preparation of Isocyanates, Carbamyl Chlorides and

Cyanuric Acid'
BY R. J. SLOCOMBE, E. HARDY,J. H. SAUNDERS
EDGAR .4ND R. L. JENKINS

At the time the major portion of this work had above 120' were also liberated by refluxing the
been completed the vapor phase reaction between corresponding monosubstituted carbamyl chlo-
phosgene and amines had not been described.2 ride in benzene or toluene.
However, a somewhat similar reaction was re- The method is applicable apparently to the use
ported recently.S The use of catalysts, preheating of any primary or secondary monoamine which
of the phosgene, and relatively long sojourn times can be boiled readily without decomposition.
(2.3-17.7 sec.) were recommended. Work re- For practical purposes the procedure is limited to
ported here more closely defines the conditions of amines containing from one to twelve carbon
the reaction and shows that the recommendations atoms. Only two conditions were found neces-
cited above are not necessary. sary for the satisfactory operation of the vapor-
This process has been made the basis of a gen- phase reaction. These were, (1) the amine vapor
eral synthetic procedure for the preparation of must be adequately mixed with phosgene, and (2)
carbamyl chlorides and isocyanates. the reaction zone must be hot enough to avoid the
The development of this procedure was separation of reaction products from the gas phase
prompted by a study of the preparation of meth- in that zone.
ylcarbamyl chloride by passing phosgene over Similarly, a t 500' ammonia was converted to
molten methylamine h y d r o c h l ~ r i d e . ~Experi-
~~ carbamyl chloride in almost quantitative yield.
ments showed that the yield of isocyanate was ad-
versely affected when the area of contact was in-
3" + COClz +HzNCOCl + HC1
creased by dispersing phosgene beneath the surface The reaction between ammonia and phosgene a t
of the molten salt. This observation suggested 400' was mentioned by Rupe and Labhard,6 but
that the reaction involved not only the liquid-gas no experimental details were given. When this
interfacial contact, but also depended upon condi- reaction was run a t 400°, considerable amounts of
tions in the gas phase. It was reasoned that some ammonium chloride were formed. In contrast, a t
of the methylamine hydrochloride was vaporized 500' this undesirable by-product was almost en-
and dissociated and that the free amine may then tirely eliminated. The carbamyl chloride was
have reacted with phosgene. If this were true, a readily converted to cyanuric acid by heating to
direct reaction between methylamine and phos- 180'
gene in the vapor-phase should be feasible. Such
3HzNCOCl- (HNC0)a + 3HC1
a reaction was carried out successfully, and meth- A consideration of the temperatures used in the
ylcarbamyl chloride was obtained, which was con- vapor-phase reaction and in the thermal decompo-
verted by more orthodox methods to the isocya- sition of monosubstituted carbamyl chloride^^,^
nate in 88% over-all yield. indicates that the isocyanate is the end-product of
The vapor-phase method for the preparation of the vapor-phase reaction a t 275' when primary
methyl isocyanate was found to have wide appli- amines are employed, but, upon cooling, the iso-
cability. A variety of mono- and disubstituted cyanate may combine with hydrogen chloride to
carbamyl chlorides and isocyanates were prepared give the carbamyl chloride.
in 70 to 90% yield by the reaction between phos- That this sequence is true was demonstrated by
gene and primary or secondary amines a t 275'. removing the hydrogen chloride from the products
RNHz +
COClz RNHCOCl HCl + of the reaction before combination with the isocya-
nate could occur. As typical examples, methyl,
RNHCOCl RNCO 4-HCl phenyl and octyl isocyanates were prepared di-
RzNH 4- COClz ---f RzNCOCl HCl + rectly without the intermediate formation of the
The isocyanate was obtained from the mono- carbamyl chlorides.
substituted carbamyl chloride by treatment with Experimental
a tertiary amine. Those isocyanates which boil General Method for the Preparation of Carbamyl Chlo-
(1) The work described in this paper was done in part under Con- rides.-The apparatus is shown in Fig. 1. The reactor
tract OEM-sr-845, recommended by the National Defense Research (K) was made of 15-mm. i. d. Pyrex tubing. The heated
Committee, between the Office of Scientific Research and Develop- portion was 15 in. long and wound with l/l&. nichrome
ment, and the Monsanto Chemical Company. A portion of this ribbon having a resistance of 15 ohms. Phosgene entered
work has been described by Jenkins and Hardy in Off. Pub. Bd. the reaction zone (M) through a 6-mm. tube (A) which
Report, P. B. 27396, "The Preparation of Isocyanates, Cyanuric was drawn out to a 1-mm. tip. Amine vapor entered the
Acid, and Decontaminant 40." reaction zone through the annular space between the phos-
(2) Saunders and Slocombe, Chcm. Reus., 45, 203 (1948). gene inlet tube and the reactor tube. Indentations were
(3) Modersohn, "Action of Phosgene on Primary and Secondary made in the walls of the reactor to impart added turbu-
Amines in the Gas Phase," Off. Pub. Bd., Report P B 707. lence t o the gases.
(4) Gattermann and Schmidt, Bcr., 20, 118 (1887).
( 5 ) Gattermann, Ann., 244, 29 (1888). (6) Rupe and Labhard, Bey., 33, 233 (1900).
May, 1950 PREPARATION
OF ISOCYANATE,
CARBAMYL
CHLORIDES
AND CYANURIC
ACID 1889

Fig. 1.-Vapor-phase reaction apparatus: A, phosgene inlet from flowmeter; B, nitrogen inlet to amine feed flask;
C, amine feed flask; D, amine flow control cock; E, calibrated tip for amine addition; F, nitrogen inlet for sweeping;
G, preheater section of vaporizer; H, flash vaporizer; I, salt-bath heated with bunsen burner; J, electrical leads t o
vaporizer heaters; K, reactor tube; L, electrical leads to reactor heater; M, reaction zone; N, product condenser; 0,
receiving flask; P, thermometer for reactor outlet temperature; Q, off-gas outlet.
The salt-bath (I) for the flash vaporizer (H) was heated Yields and physical properties are given in Table I.
to a temperature a t least 100' above the boiling point of I n spite of extreme care t o avoid the presence of moisture,
the amine being used. A weighed amount of the desired crystals frequently occurred in n-butyl and s-butyl iso-
amine, if a liquid a t room temperature, was charged into cyanates when these isocyanates were distilled a t atmos-
the amine feed flask ( C ) . Nitrogen pressure (l/* to l/1 pheric pressure. Distillation a t 20 mm. gave crystal-free
p. s. i. from a cylinder reducing valve) was applied to the products.
amine feed flask a t (B) to force the liquid to the stopcock n-Butyl isocyanate showed a marked tendency to poly-
(D) which controlled the rate of amine flow. If the amine merize in the presence of the tertiary amine hydrochloride.
was a gas a t room temperature it was fed from a cylinder An equal volume of "orthene" (technical o-dichloroben-
through an orifice-type flow meter. In such cases the zene) and of dimethylaniline were used to decompose the
amine vaporizer was not used. butylcarbamyl chloride. The final reaction mixture was
The reactor (K) was heated to a temperature of 275', cooled to freeze out the amine hydrochloride. The super-
read at ( P ) , while a slow stream of nitrogen was sweeping natant liquid was decanted and distilled with good yields.
through the system from the nitrogen inlet (F) on the side (b) Refluxing with an Inert Solvent.-The crude
of the amine preheater (G). The phosgene was turned on phenylcarbamyl chloride obtained from 50 g. of aniline
to the desired rate as shown by an orifice-type flowmeter. was dissolved in 75 ml. of toluene or benzene and refluxed
The amine addition was then begun. The maximum rate until no more hydrogen chloride was evolved. On frac-
of addition of liquid amine was estimated from the rate of tional distillation, 55 g. (86%) of phenyl isocyanate was
drop formation a t (E). A phosgene:amine molar ratio of collected, b. p. 158-162'. Other carbamyl chlorides
1.3 was used. The flow of nitrogen was used only when above butyl were converted to the isocyanate by this
the operation was started, interrupted or terminated. method. Yields and physical data are given in Table I.
The nitrogen served to prevent phosgene or hydrogen Disubstituted Carbamyl Chloride.-The general pro-
chloride froni backing into the amine vaporizing system cedure was the same as for the monosubstituted carbamyl
and forming a deposit of the urea and the amine hydro- chlorides. The reaction product was refluxed with one-
chloride. Excess phosgene was decomposed with steam. half its weight of dry toluene t o drive out dissolved phos-
The limits of the conditions for the reaction are broad. gene and hydrogen chloride. The solvent was removed
Reaction temperatures of 240-350°, sojourn times of 0.24- and the product distilled. Yields and physical data are
2.6 sec., and phosgene-amine molar ratios as low as 1.2 given in Table I.
have been used with good success. Carbamyl Chloride .-The general procedure was the
same as that used in the preparation of substituted car-
Conversion of Carbamyl Chlorides to Isocyanates bamyl chlorides, with the following exceptions. In the
(a) Use of a Tertiary Amine.-Into a 1-l., 3-necked first experiments carbamyl chloride was collected as a
-
flask was placed 150 g. of crude methylcarbamyl chloride colorless liquid. In later experiments the receiver
obtained from 49 g. (1.57 moles) of methylamine. The equipped with a reflux condenser and a mechanical stirrer,
flask was fitted with a reflux condenser, mechanical stirrer and containing 1.5 to 2 . 0 1. of orthene, was attached di-
and dropping funnel. A solution of 165 mi. (2.0 moles) rectly to the vapor-phase reactor. The reaction tempera-
of dry pyridine and 165 ml. of dry toluene was added slowly ture was 500'.
with stirring. Vigorous reaction accompanied this addi- The undiluted carbamyl chloride showed a marked tend-
tion. The resulting solution was distilled through an 18- ency to solidify below room temperature, and to lose
in., helix-packed column. The yield of methyl isocya- hydrogen chloride and polymerize on standing. When
nate, b. p. 37-39', was 79 g., 88% based on amine. the carbamyl chloride was t o be used at once in the syn-
Other isocyanates below amyl were obtained by adding thesis of cyanuric acid, orthene was put in the receiver
dimethylaniline directly to the carbamyl chloride and dis- and stirred to prevent caking of the solid. No yield of
tilling. One-fourth molar excess of amine was used. pure carbamyl chloride was determined.
1890 R. J. SLOCOMBE, AND R. L.JENKINS
EDGARE. HARDY,J. H. SAUNDERS Vol. 72

TABLE I solution was maintained a t 50'. Dry hydrogen chloride

was bubbled into the solution until a constant weight was
PREPARATION OF ISOCYANATES AND DISUBSTITUTEDreached. Correction was made for any toluene or iso-
CARBAMYL CHLORIDES cyanate which was condensed in the cold trap. The
Boiling range gain in weight of the solution was taken as the weight of
of product, Yield, hydrogen chloridz absorbed. The process was repeated
Amine OC. (uncor.) % a t 80, 95 and 105 . The data are recorded in Table 11.
Methyl 37-394'5 88
Ethyl 58-6 1a 71 TABLE 11
n-Propyl 86-88b 76 ABSORPTIONOF HYDROGEN CHLORIDEBY 100 G. OF
Isopropyl 70-75" 82 PHENYL ISOCYANATE
Allyl 83-85d 75 No solvent I n 500 ml. toluene
n-Butyl 113-116'9' 70 Temp., RNHCOCI, RNHCOC1,
OC. HC1, g. % HCl, g. %
s-Butyl 99-1010 70 50 . . . . . . . . . . . . . . . . 3 2 . 3 * 0 . 9 100
Isobutyl 100-103h 78 80 . . . . . . . . . . . . . . . . 11.7 * 0 . 5 3 8 . 2 '2
Amyl (mixed) 129-140i 70 95 11.2 * O . l 36.6 * 0 . 5 4 . 5 * 0.6 14.7 * 2
Cyclohexyl 165-168' 88 105 6 . 7 ' 0 . 1 21.9 * 0 . 2 1 . 7 * 0 . 3 5.6 * 1
Phenyl 15&162k 86 ........
n-Octyl 200-204' 81
115 4 . 0 * O . l 13.2 * 0 . 3 . . . . . . . .
n-Dodecyl 130-140 (4 mm.)' 58
125 1 . 9 ' 0 . 1 6.1'0.4 ................
Dimethyl 166-168" 90 It is apparent that condensation of phenyl isocyanate in
Di-n-butyl 245-247 81 boiling toluene will not be accompanied by the formation
Diamyl (mixed) 115-125 (3 mm.)' 83 of appreciable amounts of phenylcarbamyl chloride.
(b) NO Solvent.-The procedure described for the de-
Wurtz,Ann.chim.~hys.,[3] 42,43 (1854); n2%1.38 1; termination in toluene solution was followed, except that
G o d 0.9039.' Oliveri-Mandala and Noto. Gazz. ch n. no solvent was used. The absorption of hydrogen chlo-
itaZ., 437 I , 514 (1913). Hofmann, Ber., 15,752 (la) ; ride was observed a t 95, 105, 115 and 125'. The weight
n z o 1.3886;
~ daO, 0.8669.' Hofmann, Ann., 102, 285 of hydrogen chloride absorbed was determined by titration
(1857). Anal.? Calcd.: C, 60.58; H , 9.15. Found: of the chloride in the isocyanate-carbamyl chloride solu-
C, 60.42; H, 8.84; n Z o 1.4060; ~ d2Ol 0.880. U. S. tion.1° The data are recorded in Table 11. These results
Patent 2,326,501. 0 Anal.' Calcd.: C, 60.58; H, 9.15. are in general agreement with those of Lengfeld and Stieg-
Found: C, 60.80; H, 9.02; ~ * O D 1.3989; 0.8804. litz, who reported that phenylcarbamyl chloride dissoci-
Anschutz, Ann., 359, 202 (1908). Sharples mixed ates a t 100°.ll
amylamines were used: Custer, Ber., 12, 1328 (1879). The Direct Preparation of Phenyl Isocyanate.-The
i Skita and Rolfes, ibid., 53, 1242 (1920). Hofmann, procedure for the reaction between phosgene and aniline
ibid., 3, 653 (1870). Anal.1 Calcd.: C, 69.63; H, was the same as that used for carbamyl chlorides, with the
11.04. Found: C, 70.14; H , 11.25; n z o 1.4314; ~ d'O4 following changes. The gaseous products passed from
0.8768. Michler and Escherich, Ber., 12, 1162 (1879). the reactor into a 20-mm. Pyrex tube which was heated
Werner. J. Chem. Soc.. 115. 1010 (1919). OSharules electrically to 130'. This tube extended almost to the
mixed diamylamines weie used: Atkinson, J. C&m. bottom of a 500-1111., 3-necked flask which contained 250
Sot., 105,1290 (1914). ml. of refluxing toluene. To the flask was attached an 8-
in. bead-packed column, which was heated externally to
Cyanuric Acid.-The crude carbamyl chloride was con- maintain a temperature of 110' inside the column. A
verted to cyanuric acids by heating at about 180' for three long condenser was placed on top of the column. From
hours. The temperature rose slowly from 55 t o 180°, 50 g. (0.54 mole) of aniline there was obtained 52.3-56.8
as hydrogen chloride was evolved and polymerization g. (8249%) of phenyl isocyanate, b . p. 157-162'.
progressed. When no solvent was used the cyanuric acid Analysis of the crude reaction mixture before distillation
could not be recovered without breaking the flask. When indicated the presence of only traces of chloride.
orthene was used as a solvent the product was readily ob- The Direct Preparation of n-Octyl Isocyanate.-The
tained by filtering the resulting slurry and drying the apparatus and procedure were the same as the preceding
finely divided solid a t 200-210'. The yield was 82-89%, description for phenyl isocyanate. From 320 g. (2.48
based on ammonia. The cyanuric acid, without further moles) of n-octylamine there was obtained 339 g. (88%)
processing, was 98% pure, as indicated by a Kjeldahl of n-octyl isocyanate, b. p. 200-207'.
analysis and solubility in alkali. The isomeric cyamelid The Direct Preparation of Methyl Isocyanate .-The
is insoluble in dilute alkali.g reactor and method described for carbamyl chlorides were
Ana2.1° Calcd. for CaHsOaN,: N, 32.56. Found: N, used, with the following modifications. A 20-mm. Pyrex
31.9, 31.9. tube led from the reactor to a 3-1. flask which contained
Solubility of Hydrogen Chloride in Phenyl Isocyanate .-450 g. of dry pyridine and 250 ml. of dry toluene. The
I n order to demonstrate the lowest temperature at which connecting tube extended to within 1 in. of the surface of
phenyl isocyanate can be condensed from the reaction the liquid in the flask, and was heated electrically to 140".
gases without the formation of phenylcarbamyl chloride, An 8-in. Vigreux column was attached to one neck of the
the stability of phenylcarbamyl chloride was observed flask. On top of this column was placed an 8-in. helix-
briefly.
(a) Toluene Solution.-A -
packed column, partial take-off head, and condenser, which
1-l., 3-necked flask was was connected t o a Dry Ice-methanol trap. The pyri-
fitted with a sealed mechanical stirrer, a gas inlet tube, a dine-toluene mixture was boiled vigorously. During the
thermometer, and a condenser, which in turn was con- reaction distillate boiling a t 60-80" was collected from the
nected to a trap cooled in a Dry Ice-methanol-bath. A partial take-off head. When the reaction was completed
solution of 100 g. (0.84 mole) of phenyl isocyanate in 500 the distillation was continued until the distillate boiled at
ml. of dry toluene was placed in the flask. The toluene 112'. Redistillation of ;his crude product gave methyl
--
(7) We are indebted to the Clark Microanalytical Laboratory,
isocyanate, b. p. 37-41 . From 36 g. (1.16 moles) of
methylamine there was obtained 40 g. of methyl isocya-
Urbana, Illinois, for the analytical determinations. nate, 60.7Q/&based on amine. Essentially no methylcar-
( 8 ) Rupe and Metz, Bcr., 36, 1092 (1903). bamyl chloride was formed, since the pyridine combined
(9) Hantzsch, Hofmann and Lehmann, ibid., 38, 1013 (1906).
(10) The authors wish to thank Mr. W. B. Dunlap for this analysis. (11) Lengfeld and Stieglitz, An. Chcm. J . , 16, 71 (1894).
May, 1950 ASPARTATEAND ALANINATE OF COPPER
COMPLEXES 1891

with the hydrogen chloride at a temperature above that at finic, alicyclic and aromatic amines is described.
which methylcarbamyl chloride dissociates (93-94') .6 2. Carbamyl chloride was prepared by the re-
Summary action between phosgene and ammonia at 500'.
1. The vapor-phase reactions of primary and The carbamyl chloride was of high purity and was
secondary monoamines with an excess of phosgene readily converted to cyanuric acid by heating.
were investigated. A new method for the prepara- A ~A~~~~~~ ~ RECEIVED'~
~ ~1, 1949
~ AUGUST ~ ~
tion of mono- and disubstituted carbamyl chlo-
rides and of monoisocyanates from aliphatic, ole- (12) Original manuscript received September 24, 1948.

[CONTRIBUTION FROM THE OF CHEMISTRY, ST. LOUISUNIVERSITY]

DEPARTMENT

Polarographic, Potentiometric and Conductometric Studies on the Aspartate and

Alaninate Complexes of Copper
C. LI AND EDWARD
BY NORMAN DOODY'
As a part of a general program on complex measured by means of a Leeds and Northrup PH
formation between different metallic and amino meter, assembly number 7661. To minimize er-
acid ions2 this paper3 presents the results on the rors in calculating the aspartate concentration
aspartate and alaninate complex of copper. The from the PH, the pK2 and pK3 for aspartic acid
methods used were polarographic, potentiometric was determined a t the same ionic strength using
and conductometric. carbonate-free sodium hydroxide. A simple
Several investigators have studied the complex method for determining the pK2 and PK3 of aspar-
ions formed between cupric and alaninate ions. tic acid and equation for calculating the concen-
Boorsook and Thimann4have deduced from spec- tration of the aspartate ion will be given later in
trophotometric and electrometric data that the the paper. For alanine Keefer's value7 of 9.86 for
complex formed in alkaline solutions was CuA3-. pKz was used.
Riley and c o - w o r k e r ~using
~ * ~ a cupric ion concen- Polarographic measurements were made using a
tration cell, have concluded that a very stable Fisher Elecdropode. An H-type cell was used,
complex ion of the type CuA3- is formed; how- with a saturated calomel electrode, in a thermo-
ever, the data for alaninate solutions reported in stat kept a t 25.0'. Oxygen was removed from
the two papers do not agree. Using the dropping the solutions in the cell with a stream of hydro-
mercury electrode, Keefer' has reported the com- gen and during each run an inert atmosphere
position and stability of the alaninate complex of of hydrogen was maintained over the solution.
copper and has found the complex to be CuA2, Methyl red (0.025%) and brom cresol green
but it is desirable to extend the range of his stud- (0.12y0) were used as a maximum suppressor.
ies to lower pH values by using buffer solutions. In the case of the aspartate, all solutions were
made up to 5 X M cupric nitrate and
Experimental sufficient potassium nitrate was added to keep the
Reagent quality chemicals were used without ionic strength constant a t 1.0. In the case of ala-
further purification except for the alanine which nine, the composition of the solutions was as fol-
was once recrystallized. A standard solution of lows: 0.04 i W potassium alaninate, 3 X M
cupric nitrate was prepared by dissolving about cupric nitrate, 0.06 -51 potassium dihydrogen
0.025 mole of cupric nitrate in 250 ml. of water. phosphate plus sodium hydroxide. The reversi-
To this solution excess potassium iodide was bility of the electrode reaction was tested for each
added and the liberated iodine was titrated with a analysis by determining the slopes of plots of log
standard solution of sodium thiosulfate to the i/(& - i) against E d e . , or by determining the Val-
starch endpoint. Stock solutions of the potas- ues of (Ea/, - where Ea/, and El/, are the
sium aspartate and potassium alaninate were pre- values of E d e. a t i = (3/4)id and i = ('/&& re-
pared from the amino acids and potassium hydrox- spectively. Corrections were made for residual
ide and kept in a refrigerator until used. The PH current. Half-wave potentials were reproducible
of all solutions for the polarographic studies was to =t3 mv.
(1) Brother Edward Doody of the Congregation of Christian Potentiometric studies were made using a Cenco
Brothers. Titration-pH meter, with a glass electrode which
(2) Li and Gormley, a paper presented at the Chicago meeting of
the American Chemical Society, April 20, 1948.
had been cleaned thoroughly with dilute acid,
(3) Presented a t the 116th meeting of the American Chemical allowed to stand in distilled water for a number of
Society, Atlantic City, Sept., 1949. hours, and then calibrated immediately before
(4) Boorsook and Thimann, J . B i d . Ckcm., 98, 671 (1932). use by means of a potassium acid phthalate buf-
(5) Riley and Gallafent, J . Chem. Soc., 2029 (1931).
(6) Ferrel, Ridgion and Riley, i b i d . , 1440 (1934). (8) Kolthoff and Lingane, "Polarography," Interscience Pub.
(7) Reefer. THIS J O U R N A L , 68, 2329 (1946). lishers, Inc., New York, N. Y.,1941.