6076
GORDON E. MOORE AND RICHARD M. RADGRR
although his model is undoubtedly too simple. Poundlo has found that in order to account for the temperature dependence of the frequencies in solid 1 2 , i t is necessary to consider intermolecular bonding as proposed by Townes and D a i l e ~as , ~well as the effect of the thernial motion of the molecules as proposed by Bayer. It has been pointed 0 ~ t ~ l that - l ~ values of the coupling constant obtained from pure quadrupole spectra of solids may differ by as much as 10% from the values for the same molecule obtained from microwave investigations of the gases. Thus extreme care must be exercised in interpreting these solid state values in terms of molecular bonding. The increase in the absorption frequency observed as acetic acid is successively chlorinated roughly parallels the increases observed iIi the corresponding chlorinated niethanes." If it is assumed that the frequency change from compound to compound is due to the different electron withdrawing power of substituent groups, then by considering the trichloro compounds i t is found that the withdrawal effect increases in the order: -CH(OH)J, -CONHz, -COOH, -COCl. A longer series of substituent electron withdrawal effects can be deduced from the monochloro substituted series of compounds. The observed frequencies in this series indicates that the electron withdrawal ability increases in the order -COO-, -CONH2, -COCH,l, -COCH,Cl, -COOCzH5, COOH. These series are both in agreement with those obtained on the basis of chemical evidence alone. Further evidence for this order conies from the 110) R. V. Pound. private communication. (11) R . Livingston, J . Cham. Phys., 19, 1434 (1051)
(12) €1. C:. Dehmelt and If Kruger, L Physik, 1P9, 401 (1851). (13) E1 (; I)ehmelt, Votrrui~r~\unscha~Len, 97, 398 (19.50)
Vol. 74
shifts of the carbonyl stretching frequencies in the infrared spectra of these compounds. The carbonyl stretching frequency is presumably related in some way to the charge on the carbon atom of this group. The same residual charge determines the electron withdrawal effect in these series of compounds, hence one might expect a correlation between the observed quadrupole coupling constants and the frequency of the carbonyl stretching frequency in these compounds. Unfortunately these frequencies are not available for all these chlorinated compounds, but if we consider the analogous non-chlorimted compounds14 i t is found that the C-=O infrared frequencies increase in the same order as the coupling constants. The carbonyl frequency of chloroacetone could not be found in the literature; however, a measurement shows this frequency to be slightly higher in chloroacetone than in acetone as might have been predicted. From this evidence i t would seem that either the solid state effects are relatively constant in this series or they do not differ enough to invert the order of these series. The correlation of solid-state quadrupole-coupling constants with Hammett's u for a series of substituted benzenesI6 seems to lend additional support to this hypothesis. Acknowledgment.-The author wishes to thank Professor E. Bright Wilson, Jr., for many helpful discussions during the course of this work. He is also indebted to Messrs. L. Hedrick, H. Meal, G. Jones and C. Dean for extensive help on the instrumentation. (14) Iiandell, Fowler, Puson and Dangl, "Infrared Determination o f Organic Structures." D. Van Nustrand Co., Inc., N e w York, N. Y.. I !+I!J. (1.5) 11. C. Meul and IS. B. Wilson, J r . , private communication.
CANBRIDGE 38, M A S S .
[CONTRIBUIION FROM CALIFORNIA INSTITUTE O F TECHNOLOGY]
The Infrared Spectra and Structure of the Chloramines and Nitrogen Trichloride' BY CORDON E. MOOREAND RICHARD M. BADGER RECEIVED J U N E 9, 1952 The infrared vibration-rotation spectra of gaseous NH2C1, NHDCl, NDCL and NCla were investigated from 1.4to 25p. Several fundamental vibrations were identified, and the large rotational constants were evaluated for NHZC1, NHDCl and NHClz. With the assistance of reasonable assumptions regarding other paramete5s these were used t o calculate an H-N-Cl angle of 102' in NH,Cl, and suggest that LCl-N-Cl = 106" and Q - C I = 1.76 A . in NHC12. These parameters are interpreted on the basis of simple electronic considerations.
Introduction The chloramines may be regarded as ammonia with one or more of the hydrogen atoms substituted by chlorine. While they have been known for a long time,* only the completely substituted compound NC13 has been investigated extensively. However, as far as we are aware, even its structural parameters have never been determined. Information concerning the properties of dichloramine is particularly lacking. The reason for this lack of experimental data probably lies in the (1) Contribution No It396 from tlir G a t e s a n d Crellin Lahoratories
of Chemistry. (2) For a good review of their chemi\try see J. 1; T. Berliner. J A rn. W o l e r Works Assoc., 19, 1320 (1931).
instability of these compounds. Both monoand dichloramine decompose readily to yield, among other products, the sensitive and powerful explosive nitrogen trichloride. Since these molecules presumably have a rather simple structure, i t was felt that a study of their infrared vibration-rotation spectra should allow one to draw some conclusions regarding their molecular configuration. Experimental Samples were prepared by reaction of aqueous NHs and NaOCI at 0'. The resulting solution was warmed t o 1525' and connected through a CaCL filled drying tower to an evacuated cell. After nearly every preparation it was necessary to cleanse the entire system and replace the desic-
SPECTRA AND STRUCTURE OF CHLORAMINES AND NITROGEN TRICHLORIDE 6077
Dec. 5, 1952
TABLE I Cm. - 1
I5
NHECl 6522.9 4893.8 3380.0 2020 1553 1032
686
M M S W
S VS VW
VI
vi
VIBRATIONAL FREQUENCIES OBSERVEDFOR THE CHLORAMINES Assignment Cm. -1 I" Assignment Cm.-I NHClz 4- ub (A*) -t V I (A")
(a")(assym N-H stretch) 2ua (A') Y % (a')(NHz "scissors" bend) ua (a')(HzN-CI bend) Y , (a')(N-CI stretch) vi
6393.9 3279.0 2584 1960 1295 1002 687b
666b
M
2v1 (A')
VS W W M VS
(a')(N-H stretch) 2vs (A') 2va (A') Y L (a")( H bend out of plane) VI (a')(H bend in plane) YO (a') (sym. N-CI stretch) YO (a")(asym. N-CI stretch)
S S
VI
'[
1
a Intensity: VS, very strong; S , strong; M, medium; W, weak; VW, very weak. Absorption increasing a t 400 cm.-l cut off of KBr prism. centers.
1273 1021 652
-390'
Assignment
15
NCla W W S ?
2(N-CI stretch) (Stretch bend) N-CI stretch Bend
+
NHDCl 3339.1 2490
S M
(N-H stretch) ( N - D stretch)
* Maxima, but probably not band
cant since traces of products from previous preparations ably accurate to 0.3 ern.-'. Relative frequencies of closely appeared to act as catalysts for the decomposition. The adjacent lines should be considerably more accurate. gas collected in the cell contained some mixture of "3, Results NH2C1, NHClz and NCl, as well as some Nz and occasionally a little NzO. The exact composition of the mixture deThe prismatic spectra obtained for "$21, pended on the ratio of NH3 t o NaOCl and the pH of the NHClz and NCla reduced to per cent. absorption, solution. In general more NaOCl and lower pH favored the are shown in Fig. 1. Bands observed with the more highly chlorinated products. First attempts to prepare half-deuterated chloramine by distilling from a solution containing about 50% heavy water were completely unsuccessful. The chloramine and ammonia collecting in the absorption cell were undeuterated. This was probably due to a very rapid exchange between hydrogen in the chloramine and ammonia and that in water remaining in the desiccant. This rapid exchange proved to offer a very effective method for preparing heavy chloramine. In the second method the desiccant was thoroughly baked out under vacuum and a few tenths of a milliliter of D2O (99.8%) were added to the top of the drying column and heated to disperse it throughout the CaC12. An ordinary mixture of ammonia and hypochlorous acid was then distilled through the deuterium treated CaClt and the gases collected in an absorption cell in the usual manner. Good exchange took place and the collected gases were rich in deuterated compounds. Since our primary interest was in the half-deuterated monochloramine NHDCl, only a small quantity of heavy water was employed. Mixtures of ammonia and monochloramine were quite stable and could be investigated for about two hours before decomposition became troublesome. Unfortunately the ,'" bands of these two compounds strongly overlap one another, :-2- :2 making measurements difficult. Samples containing only 2 4 8 8 10 12 14 16 NHzCl or a mixture of NHzCl and NHClz decomposed autoIVave length in microns. catalytically shortly after their preparation. For a typical sample decomposition was first noticed seven minutes after Fig. 1 -Prismatic spectra of NHzC1, NHCl, and NCh preparation and 45 seconds later was essentially complete. reduced to per cent. absorption: path lengths 80 cm.; total No such sample persisted more than 20 minutes. This decomposition necessitated observing several bands in sec- pressure, 10 cm. in each case; partial pressures unknown. tions. Samples containing the most NHCh were the shortgrating as well as one prismatic band near 10 est lived. We were not able to observe the spectrum of NHClz without considerable NHlCl present. While am- are shown in Figs. 2-6. The frequencies of all monia inhibits the decomposition of ?rTHzC1,it seems to react rapidly with NHCls. Accordingly no samples contain- band centers observed for these molecules together ing NHCl? in the presence of NHa were obtained. Nitrogen with their vibrational assignment appear in Table trichloride was observed primarily as a decomposition prod- I. Measured frequencies of sub-bands for several uct from the samples rich in NHC12. Samples of NC13 de- bands of NHzCl and one band of NHDCl and the composed quite slowly and could be investigated for well corresponding rotational assignments are given in over an hour. Three spectrographs were employed to investigate the Table 11. Similar information for one band of spectra. A Beckman IR-2 spectrophotometer was used in NHClz appears in Table 111. the rock salt region. A vacuum prism instrument with Vibrational Assignments KBr optics was used from 12 to 25 p . 3 The region from 1.4 to 3.2 was investigated under high dispersion with a vacA non-planar tetratomic molecule of point group uum grating spectrograph4 employing a 7500 lines/inch replica grating. Lines in the 1.4 and 1.9 p water bands6 G(NH2Cl, NHC12) has four fundamental vibrations symmetric with respect to the symmetry plane in the first and second orders were used for calibration. Frequencies measured with the grating instrument are prob- of the molecule (a') and two anti-symmetric with
INC" i I
(3) This instrument was a gift from the Shell Development ComFor a description see R. R . Brattain, P h y s . Rev., 60, 164 (144 1). (4) R . h l . Badger, I. H . Zumwalt and P. A. Giguere. Rev. Sci. znslr,rrmCnfs, 19, 861 (1948). (5) R . C. Nelson. Summary Repnrt No. I V , Contract NObs 28373,
pany.
Dept. Physics, Northwestern University.
;i
, d
respect to this plane (a"). For the molecules under consideration these will be designated as in Table I. NH2Cl.--The bands of "$21 should be of two types which should be more or less easily distinguishable. The a " fundamentals should be of
Fig. 2.--u6 of NHsC1: v I of ?iI-IC12s h o w weakly a t low frequency end of tracing; path length 80 cm.; total pressure, 10 cin.; niaximurn absorption is about 4sr;,
I F
B
8
1200
1000
IIOO
9OOCu;'.
Fig. 3,--v3 of SHpC1: weaker maxima are probably ~ ~ ~ path 1 3 7length ; 80 cm.; total pressure, 10 UTI.; iriasiinuin ahsorption is about 8
