Hydrogen-bonded complexes of iso-cyanic acid: Infrared spectra and thermodynamic measurements

S~~~trochimic8 Aota,Vol. 26A.pp. 109to 120. PergamonPress1970.Printedin NorthernIreland

Hydrogen-bondedcomplexes of iso-cyanic acid: Infrared spectra and thermodynamic measurements JANE NELSON Department of Chemistry. Queen’s University, Belfast (ReceCued16 January 1969) Abstract-Changes in the infrsxed spectrum of iso-cyanic acid on hydrogen-bonding am discussed. Free energy, enthalpy and entropy changes for the 1: 1 association of iso-cytmic acid with a number of bcscs in carbon tetrschloride solution at 25’C srs reported. The results are discussed with reference to similar results for iso-thiocysnic acid.

THERE is little thermodynamic data available on the hydrogen-bonding properties of N-H acids. In this laboratory, we have carried out a comprehensive investigation of the complexes formed by iso-thiocyanic acid with a wide variety of Lewis bases [l-3]. The present work extends the investigation to the hydrogen-bonding properties of iso-cyanic acid, HNCO. It is of interest to discover whether, in view of the structural similarity of this molecule to HNCS, it has comparably strong protondonating tendencies. Previous investigations of HNCO have been mainly concerned with its spectra and structure. There have been no previous investigations of its proton-donating properties. EXPERIMENTAL

Iso-cyanic acid, HNCO, was prepared by depolymerisation of cyanuric acid at 600°C in ‘uwuo. The HNCO produced was collected at - 180°C and solvent was added to the solid at this temperature, under nitrogen. The solution was then allowed to warm to room temperature under nitrogen. Concentrations were found by interpolation on a calibration graph relating concentration (found by titration against aqueous sodium hydroxide or by the Kjeldahl method) to optical density in an O-2 mm path length cell. Rapid polymerisation of HNCO occurred in concentrated solution, but the dilute (1O-2-1O-3 M) HNCO solutions used for determination of equilibrium constants showed no deterioration over a period of two to three days. The bases used were the best grades obtainable commercially. Pyridine was dried over potassium hydroxide pellets followed by molecular sieves. The ethers were dried over sodium wire followed by molecular sieves (for propylene oxide, molecular sieves only). Acetonitrile was dried by refluxing with successive portions of phosphoric [l] T. M. BARAKAT.N. LEWE and A. D. E. PULLIN,Trans. l%ra&zy Sot. 59, 1764 (1963); 59, 1173 (1963). [2] T. M. BAFUEA T, 8. M. NELSON,J. NELSON snd A. D. E. PKLTXN,Tram. Faraday Sot. 62, 2674 (1966). [3] T. M. BARAKAT, S. M. NEILSON, J. NELSON and A. D. E. PXJLLIN,Tram. Faraday Sot. 65,41 (1969). 109

110

JANE i%ELSON

oxide. In all cues final purification was by fractional distillation and purity was ohecked by vapour phase ohromatography. Hexaethylbenzene (Aldrich) had the literature melting-point of 164~5165&'C and was used without further purification. Equilibrium quotients for association in carbon tetrachloride solution were determined, as described previously [2],from the intensities of the N-H absorption at 3473 cm-l in unbended HNCO, using a Perk&Elmer Model 21 spectrometer equipped with LiF optics and a Research and Industrial Instruments Co variabletemperature cell fittedwithInfrasi1 windows. Enthalpy and entropy changes (strictly, apparent enthalpies and entropies of association as concentrations were used instead of activities) were calculated from the equilibrium quotients at -20, 0, 20, 4O“C. Each equilibrium quotient is the average of four to seven separate determinations. RESULTSAND DISUUSSION The infrared spectrum of HNCO has been studied in the gae-phme by REID and HEBZBER~[4] and by MOBCJAN and LAWSON[S]. ASHBY and WEB~XR [6] have re-investigated the spectrum below 1000 cm- l. These workers’ assignments are given in Table 1. Table

1. Assignments

(cm-l)

va

VI)

*4

Vti

3531 3620

2274 2271

1327 1342

572

670

3473 3127 2849

2265 2249 2244

1321 1310 1306

797 572 761 660 786 -860 ~920

670 660 577 ?680 -

(797) 677 777 7670 ?666 9666

Vl

REID end HEIUFIEB~ [4] MOBC+ANand LAWSON [B] Akzm~and wERNEB[6] DIXON and KIRKBY [7] HNCO in inert 6olvents HNCO diethyl ether complexes HNCO/pyridine complex

for HNCO v4

There is no question of the assignment of the three highest energy vibrations to v,, the N-H stretch, Y*,the pseudo-antisymmetric NC0 stretch and Y,, the pseudosymmetric NC0 stretch, in that order. In inert solvents these frequencies are not much altered from the vapour-phase values. The solvent shift of y1 of HNCO in carbon tetrachloride solution is 58 cm-l, only 9 cm-l less than that for HNCS [l]. Shifts in the N-H frequency on hydrogen-bonding are considerably smaller for HNCO than for HNCS [2,3].The hydrogen-bond shifts, AvN_u, for both HNCO and HNCS complexed to Lewis baaea are given in Table 2. It will be noticed that the basicity order, m measured by AvN_u, is the same with HNCO as with HNCS (see Fig. 1). The plot of AY, (HNCS/base) against AY, (HNCO/baee) is linear within error limits, suggesting that the mode of interaction of the two acids with any particular base is similar. The half-band width of y1 increases from 32 cm-1 in the free acid to 600 cm-l in the HNCO/pyridine complex. Generally, the larger the shift, the broader the band. Consequently, y1 (bonded) is many times more intense than y1 (free) as is general in H-bonding acids. Figure 2 illustrates the relationship between Avlla and Av for a [4] L. REID

and G. E~RZBERC~, Dieucw&w~ Family Soo. MO, 92 (1950). [6] H. W. MORGAN and J. R. LAWSON, Co2u~SpwtrySy~p. (1961); privati communication. [S] R. H. ASEBY snd R. L. WERNXR, J. Mol. &uecby 18,184(1966). [7] R. N. DIXON and G. H. KIRKBY, Traw. Faraday Sot. 64,2002 (1968).

Hydrogen-bonded complexes of iso-oyenic acid

111

Table 2. Hydrogen bonding shifts of v, in HNCS mndHNCO B&%3 Propylene oxide Tetrahydmfumn Et,0 Il-PQO Tetmbydropymn iao-Pr,o 1,,4-Dioxme Dibemyl ether Anisole Me&ylene Nitrobenzene Diphenyl ether Aoetonitrile Xl-BU*S t-B@ Hexmthylbenzene Hexsmethylbenzene Nitrometham

Avl (HNCS)

Avl (ENCO)

(cm-l)

(cm-l)

290 660 630 664 646 668 448 440 360 124 166 237 280 400 440 170 110 146

f f f f f f * f f & * f f f f * f f

266 f 361 f 346 f 362 f 367 -& 367 f 298 f 283 f 206 f 83 f 98f 132 f 180 f 286 f 318 f 133* 130 f 83 f

20 30 30 30 30 30 30 30 20 10 10 10 16 30 30 10 10 10

10 16 16 16 16 16 16 16 10 7 7 7 10 16 16 7 7 7

AvIp-Hin aoetonitrile oomplexe43 extrapolated to iniinite dilution.

0 Ethers 0 Sulphidas x Nitrilcs + Aromotics 6lx

.

0 Nltromethmc

4oc

T

5

9 x &xx Y I

-I

d eci

ICK

loo

/

,

,

200

300

400

AY,_,(HNCO/IXS(I), cm-1

Fig. 1. Shifts in vl of HNCS and HNCO on hydrogen-bonding.

112

JANN NELSON

6CC-

I 200

I 400

AU

$2’

I 600

cm“

Fig. 2. Relation between the shift on hydrogen-bondingand the half-width of the bonded absorption.

series of Lewis bases bonded to HNCS, HNCO and HN,. This is satisfactorily linear considering the error in measuring both quantities in these broad and sometimes irregular (because of base absorption) bands. Ye of HNCO in carbon tetrachloride solution has moved 15-20 cm-l to lower frequency from the gas-phase value. Ye in HNCO is a strong, sharp band (AY~,~= 11 cm-i) which is broadened on bonding, having a half-width of 20 cm-l in the HNCO/diethyl ether complex and of 40 cm-l in the HNCO/pyridine complex. The area of the band is only slightly increased however. On hydrogen-bonding, the maximum moves slightly to lower frequencies but never sufficiently to be resolvable from the absorption due to the free acid. Even with pyridine, which has a Av, shift of 600 cm-l, Av, is only about 11 cm-l. In HNCS, the pseudo-antisymmetric stretch is much more sensitive to hydrogenbonding. Shifts of ~40 cm-l are recorded on bonding to ethers, where Av, is around 540 cm-l. It is interesting that the shift in v2 on hydrogen-bonding is towards lower frequencies in HNCO, whereas the much larger shift in v2 of HNCS on hydrogenbonding is toward higher frequencies. This reflects the participation of ionic structures in the hydrogen-bonded complex, as the frequency of v,(NCS-) is higher, and v&NCO-) lower, than in the unionised acids. It should be noted that polarisation towards a H-+Nz&-Xstructure, which is expected to be appreciable, at least in the case of HNCS [3], will reinforce the shift towards the ionic frequency. v~, the pseudo-symmetric stretch, is of low intensity due to the small change in dipole moment in this mode. One might expect the electronic symmetry of the N=C=O group to be disturbed on hydrogen bonding, resulting in intensifkation of v~,and this is in fact observed (Fig. 3). Accurate area measurements are not possible

Hydrogen-bonded

0

-

C

_

aomplexea of iso-cyenio aoid

113

-

Frequency,

wovenumbers

Fig. 3. Intensification of Y* of HNCO on hydrogen-bonding Curve 0: WC0 in CS, solution. Curve 1: HNCO/propylene oxide complex in CS,. Curve 2: HNCO/trimethylene oxide complex in CS,. Curve 3: HNCO/pyridine oomplex in CS,. Curve 4: HNCO/phenylacetonitrile complex in CS,. All solutions have the same HNCO concentration.

because of displaced proton-acceptor bands appearing in the wings of vs, but a rough estimate indicate that vais intensified by a factor of approximately 3 where the proton-acceptor ia benzyl cyanide, and by 4,5 and 12, where the proton acceptors are propylene oxide, trimethylene oxide and pyridine, respectively. A displacement of Yeto lower frequency accompanies the intensification. Shifts in va (where this absorption is not obscured by displaced acceptor bands) are given in Table 3. It can be seen that complexing to pyridine produces a shift of about 15 cm-l, and to ethers a shift of around 10 cm-l, while the frequency is scarcely affected by complexing to nitriles or sulphides, although the band is broadened and intensified. In the spectrum of HNCO in the gas phase there remain three bands to be assigned to Ye,Y, and ye. There has been some controversy over the assignment of these modes. Because of Coriolis coupling between ~a and the low-lying a1 vibrations, no pure perpendicular band is observed in the gas phase. Also, the results of shifts on deuterium substitution depend on correction for Coriolis interaction before they can be used 8

114

JANIUNELSON

to indicate which of the modes is vg. ASHBY and WERNER[6] have discussed the Coriolis perturbations of the three lowest frequencies of HNCO, and have evaluated Coriolis coupling constants Esa2and Eds2 However, a recent communication [7] points out that these authors have applied the Coriolis correction with the wrong sign to the 577 cm-l band. This invalidates their assignment of ve to the 677 cm-l band on the basis of the similarity of the frequencies in the deuterated [8] and undeuterated molecules. DIXONand KIRBY [7] prefer the assignment of the 777 cm-l band of HNCO to vs. This is also the suggestion of MORGANand LAWSON[5] made on the basis of a normal co-ordinate analysis. A study of the shifts in the three lowest frequencies of HNCO on hydrogenbonding should be helpful in deciding which of these vibrations involve hydrogen Table 3. Shifts in v8 of HNCO on H-bonding B&Xl3 None Et20 Pyridine Trimethylene oxide * Propylene oxide Tetrahydrofursn * CH,CN * PhCH,CN * t-Bu,S

*v2

v2 (cm-l)

(cm-l)

1321 1310 1306 1311 1316 1309 1318 1318 1315

11 f3 16 f3 10 f 3 6f4 12 *3 3*4 3i4 6zt4

-

* Shifts for these low K systems obtained by graphical separation of free and bonded absorptions.

bending motion. Unbounded HNCO in carbon disulphide solution shows a very broad absorption with maximum at 785 cm-l. This has a broad shoulder, which on graphical separation is found to have a maximum at 680 cm-l. There is also a fairly broad, though well-defined band at 670 cm- l. Because of the appearance of strong displaced acceptor bands in this region, the analysis of these broad maxima is made difficult in the complexes, but it is clear that the general effect is to move the 785 cm-l band to higher frequencies (by up to 140 cm-l, Fig. 4). This agrees with the assignment of this mode as the N-H in-plane bend as suggested by Herzberg and Werner. It does not seem likely that this absorption corresponds to vs, which involves no hydrogen motion. Another indication that the ~780 cm-l band is not vg comes from the sign of the (A’ - A”) term for the 659 cm-l band of HNCO. Ashby and Werner have found this to be positive, indicating Coriolis coupling to a band at lower fiequency. However, if the ~780 cm-l band is not vg, then it appears that the product rule is not obeyed [23) for ve of HNCO. Because of the complex nature of the spectra in HNCO/base systems, it was not possible to ascertain the position of the broad low frequency shoulder of the 785 cm-l band in the complex. The 670 cm-l band is unaffected by hydrogen-bonding, which [S] R. H. ASEBY and R. L. W-,

&.wctrooMm.Aota aZ, 1345 (1906).

Hydrogen-bonded oomplexes of iso-cyanic acid

116

suggests it has little hydrogen deformation character. A very broad new band is seen in I/IWO/ether complexes, but not in the HNCO/pyridine complex, in the 400-500 cm-l region. C%angesin the proton-acceptor .3pectmtm Hydrogen bonding is known to produce small changes in the proton-acceptor spectrum, presumably reflecting perturbation of the electronic distribution on complex-formation. In the HNCO/ether complexes, bands due to the ether moiety of

0.2 -

IO. I low

I 900 Frequency,

I 900

I 700

wavenumbers

Fig. 4. Changes in the spectrum of HNCO in the 1100-600 cm-l region on Hbonding. Curve 1: HNCO in CS, solution. Curve 2: HNCO + excess trimethylene oxide in CS,. Curve 3: HNCO + excess pyridiue in CS,. (The three solutions have the came HNCO conoentmtion.)

the complex appear when the free ether is compensated out, at frequencies 10-15 cm-l lower than the corresponding free ether absorptions (see curve 2 in Fig. 4). Because of the relatively small association constants, a considerable excess of ether was necessary in order to produce an appreciable concentration of complex, and doubling of the ether peaks due to absorption by free and complexed ether was not observed; accurate frequencies for the displaced ether bands were thus not obtained Similarly, for the HNCO/acetonitrile complex, displaced nitrile absorption was observed in the compensated spectrum on the high frequency side of free v,+x, but an accurate value for the Avc, on hydrogen-bond formation could not be obtained. The room-temperature equilibrium constants for the association of HNCO with pyridine were high enough to permit observation of doubling in the spectrum, uncompensated for base, due to the presence of free and complexed pyridine. In the

116

JAXE

NELSON

HN~O/p~i~e complex the intense symmet~c pyridine ring vibration at 992 cm-1 (YJ‘shifts to higher energy by 12 cm-l and other vibrations in this region are shifted to higher frequency by 3-6 cm-l on complexing (see Fig. 4, curve 3). Shifts to high frequency of many pyridine vibrations are observed when pyridine acts as a Lewis base, e.g. in pyridine/halogen complexes or in pyridinium ion or the coordination compounds of pyridine with transition metals. The y1 frequencies for pyridine in different acid/base complexes are given below [16] : Compound y1

pyridine

pyridine/ICN

pyridinium ion

992

1003

1010

(cm-l)

Pt.Py&l,.ZH,O 1018

These changes are the result of the positive charge acquired by the pyridine nitrogen in the complex, which modified the electron distribution in the n-orbitals, No changes in the spectrum of ditertiarybutyl sulphide or hexaethylbenzene complexed to HNCO were observed, but for these bases only a small proportion was in the complexed form at room temperature. T~~~d~~~~c

rn~a~~~~rne~~~

Thermodynamic functions for the association of HNCO with a base B: HNCO + B + B.HNCO in carbon tetrachloride solution at 25°C are given in Table 4, together with Arr, the shift in the ‘N-H stretching mode on hydrogen bonding. Statistical errors as calculated in Ref. 191 are ~~ca~d. Table 4. Thermodynamic data for HNCO complexes

Base Diethyl ether Propylene oxide Tetrahydrofunwn. Acetonitrile Ditertiary butyl sulphide Pyridine Hexaethyl benzene

--A%

-wxi

--dsz,

rr, (mo1iFf-l1.)

(koal. mol&)

(kcal. mole-l)

(al. deg-’ mok+)

6.68 p@27 4.76 ~0.26 9.10 PO.06 4.60 rp@4 1.43 ~0.07 33.9 PO.07 1.36 ~0.05

l-12 0.93 1.31 0.90 o-21 2.09 0.18

3*92 po*os 3.49 ~0.14 4.24 ~0.08 3.21pO.08 2.84 ~0.08 6.42 ~0.14 2*41~0*24

9.4 8.6 9.8 7.7 8.8 11.2 7.5

A91

(cm-l) 346 226 361 182 318 624 133

+ f & + & f f

15 16 15 16 16 40 10

p represents90% con&dencelimits 88 determinedin Ref. [9].

The enthalpy of association of any particular base with HNCO is approximately two-thuds of its enthalpy of association with HNCS. Figure 5 shows the relation between AH,,,,/HNCO and AlY,,,,,/HNCS, which is linear within error limits. The slope of the line is the same, within error limits, as that of the ~~~~~~~~~~~~ vs. lot (Fig. 1). Figure 5. confifms the inference of Fig_ 1 that HNCO is an A~nxoo~aeej P appreciably weaker proton-donor than HNCS. This is also demonstrated by the formation of soluble hydrogen-bonded complexes of isocyanic acid with pyridine and propylene oxide. Thiocyanic acid rapidly attacks the epoxide ring in propylene oxide [9] LAITENEN,

Chenz~cuEAlaalysis, p. 457. McGraw-Hill (1960).

Hydrogen-bonded complexes of iso-cyanic acid

117

to give a hydroxy-thiocyanate [2]. With pyridine in carbon tetrachloride solution, thiocyanic acid reacts with proton transfer to give a precipitate of pyridinium thiocyanate. AH/ Av, correlation It was observed [2, 31 when HNCS was used as proton-donor that no universal AHlAv correlation could be obtained where different types of proton-acceptor were employed. A satisfactory AHlAv correlation could be obtained, however, for a series of closely related proton-acceptors such as ethers. The AH/Av points for the association of HNCO with diethyl ether and tetrahydrofuran fit the correlation for the

-AH

HNCS I bore’

kcal mole-’

Fig. 6. Relation between the enthalpy of hydrogen-bonding of a brtseto HNCO and FfNCS. (Symbols as for Fig. 1.)

the association of HNCS with ethers [2] though the point for the strained ring ether, propylene oxide, is well off the line, presumably because of the change of hybridisation of the oxygen atom, from 8~9 is unstrained ethers to spa [lo] in this compound; propylene oxide might be expected to behave more like a carbonyl compound than an ether. WEST’Sresults show [ll] that ketones deviate from the AH/Av correlation for the association of phenol with ethers in the direction of too high AH. Figure 6 shows the correlation between AH and Av for association of HNCO with the seven bases in Table 4. This demonstrates that Av can only be taken as a qualitative guide to AH. For polar compounds such as nitriles, and for aromatics, the process which controls the lowering of AvN_n accounts for only a part of the total enthalpy of association. However, the same pattern of AH/Av relationships is observed as with HNCS. The best line through the data passes close to the ether points, acetonitrile and hexaethylbenzene deviate in the direction of too high AH, and [lo] E. LIPPERT and H. PRIQGE, Ber. Bumengm. Phys. C?mn. 67,415 (1963). [ll] R. WEST ad D. POWELL,Spectrochim. Actcc 20, 983 (1964).

Fig. 0. Correlation between the enthalpy of aaeociation and the shift on hvdrojzen-bonding for

I

I

I

200

400

600

Av,_,

,

cm”’

ditertiarybutyl sulphide in the direction of too low AH. Pyridine (which could not be studied with IINS) is well off the line in the direction of low AH. Deviations in this sense have been observed previously 1121for the association of pyridine with alcohols (compared with alcohol/ether and alcohol/carbonyl complexes). An explanation of the different AHjAv, behaviour of different types of protonacceptors has been suggested for HNGS. Resonance st~ctur~ (2) and (3) below are assumed to account for the lowering of vi: B H-NCX

(2) B II+ NCX

(3) B+--H NcX

(4) B H-&C--X-

(1)

If structures other than (2) and (3) are important in describing binding in the complex, Av, will underestimate AH. In the case of polar bases such as nitriles it is suggested that structure (4) is important, leading to Iow A9 relative to AH, whereas for pyridine there is evidence from P&El% studies that (3) is important, leading to high A* relative to AH. The PMR, signal for HNCO bonded to pyridine is narrower by a factor of 10 than for unbended HNCO or HNCO bonded to other bases. This indicates a rapid exchange process* :

* A largeinereaaein the rate of nitrogen quadrupolerelaxation in the pyridinecomplex could ttleo explain the sharp singlet observed, but broad triplet eignals are observed for the ether oomplexea. The rate of relaxation would not be expeoted to be vet differentin HNCO/pyridine and HNCJO~ether oomplexes, [lZ] E. D. BECKER,~pe~~~~~.

A&a I?, 436 (1901).

Hydrogen-bonded

oomplexes of iso-cyanio acid

119

The smaller than normal Avl shift for aromatics oannot be attributed to polarisation effects, but there may be some additional interaction between the r-system of HNCS and the aromatic n-electron system which accounts for the unexpectedly large AH relative to Av for aromatic hydrocarbons. YOSTIDA and OSAWA [13] have suggested that interaction of the charge transfer type is important in the hydrogen-bond formed between phenol and v-bases. OKI and MUTAI [14] have demonstrated from the relative proportions of the rotamers of p-nitrobenzyl phenylamine that intramolecular charge transfer between two P-systems can have an energy comparable to that of an N-H - * + VThydrogen bond. AHjA# correlation In the absence of large differences in solvation or steric effects, a roughly monotonic AH/AL? relationship is to be expected [3], as observed for the HNCS/base complexes. A fair correlation is obtained for the HNCO complexes studied. In fact the HNCO points can be included in the AH/AS diagram for HNCS complexes without increasing the scatter [3]. The chief deviation observed is with the ditertiarybutyl sulphide complex, which is off the line in the direction of high -AL?. It must be remembered, that the low association constant in this case necessitated the use of 0.6 M ditertiarybutyl sulphide in carbon tetrachloride as solvent medium, and it is possible that solvation effects are predominant.

C0NCLm310N

It is clear from this work that HNCO is a much weaker proton-donor than HNCS. This is also the inference of the dissociation constants of the acids in aqueous solution: for HNCS pK, = -1.38 [17] and for HNCO pK= = 3.66 [18]. However phenol [ 1S] with an acid pK, of 10 is intermediate in strength as a hydrogen-bonding acid, so there is no general correlation between pK, and proton-donating tendency, as measured by the enthalpy of association with bases in inert solvents. This is not surprising in view of the importance of solvation in aqueous ionisation equilibria. Another indication that HNCO is a weaker proton-donor than HNCS comes [16] from NMR data. The chemical shift of the proton in HNCO (extrapolated to 6.66 T at infinite dilution in Ccl,) is at higher field than that in HNCS (5.06 7 at infinite dilution in Ccl,) and this is unlikely to be entirely accounted for by differences in the anisotropic shielding since the two molecules have similar N-C bond orders [21]. The lower shielding of HNCS probably reflects lower electron density near the proton. Structural data [19] are available for the acids, but they do not provide any [13] YOSEKIDA and E. OSAWA,J. Am. Chern.Soo. 88.4019 (1966). [14] M. 0x1 and K. MTJTAI, TetrahedronLett. 2019 (1968). [15] J. NELSON, Ph.D. Thesis, Belfast (1968). [lS] I. HAGUE and J. L. WOOD, Spectmch&n.Acta 23A, 959 (1967). [17] T. D. B. MORGAN, G. STEDE~, P. A. E. WHINCUF, J. Ckmn. Soo. 4813 (1965). [US] H. FREISER and Q. FERNANDO, IO& Eqdibria in Analytical Chemistry. [IQ] R. KEWLEY, K. V. L. N. SASTRY and M. WINNEWISSER, J. Mol. Spectry 10,418 (1963).

JANE &!lLSON

120

obvious explanation of the difference in proton-donating power:

rN-.[lgl HNCO HNCS

0.987 f 0.01 0.988 f 0.003

HNC 128%’ f 30’ 134%9 * 10’

6.90 7.00

The N-H bond length is considerably shorter than in ammonia derivatives, and the bond angle at the nitrogen considerably greater [20],indicating a larger amount of s-character in the acids than in amines. The bond length is the same, within experimental error, in the two acids, but there is a difference in bond angle well outside the error limits indicated. This suggests that HNCO has slightly less s-character in the N-H bond than HNCS. The resonance form H-&---X would then be less important in HNCO, and acidity, consequently, slightly lower. Table 6. Results of Wagner’s calculationsfor HNCO, HNCS and their conjugate ions Atomic charges

HNCO NCOHNCS NCS-

QN

Qo

Qx

O-2908

-0.0372 - 0.0442 0.6709 0.1934

-0.2636 -0.1846 -0.5134 -0.7108

-0.7712 -0.1575 -0.4826

a-bond orders P AB P BC 1.6917 16413 1.5001 1.8243

1.2104 1.1422 I.1400 0.7964

dSc0

32.6793 33.7125 29.7209 31.9985

a-energy 1.1332 2.2776

Another important factor which must be considered in examining protondonating power is the difference in r-electron energies between the acid and its ion. The n-electron energies for theisocyanate and iso-thiocyanate ions and their hydracids have been calculated [21] by WAGNER, using a L.C.G.O. (unsymmetrical group orbitala) method. In the pseudohalide anions, both N-C and C-X bonds are quite strong, but in the hydracids N-C is weakened and C-X strengthened relative to the conjugate ions. This means that the acid strength is not simply proportional to the charge on the nitrogen atom, because of the difference in bond characters. Wagner finds (Table 6) the 7r-stabilisationenergy on forming NC& from HNCS to be twice that gained on forming NCO- from HNCO. This effect, rather than charge distribution in the molecule, appears to be predominant in determining the relative hydrogen-bond donor strengths of the acids. Achmowledgemem+-Theauthor wishesto thank Professor N. SHEPPARD for reading this manuscript and for helpful comments. Thanks are also due to the Ministry of Education of N. Ireland for a post-graduate studentship and to Queen’s University for a Foundation Studentship. [20] [21] [22] [23]

Chem. Sot. Spec. Publ., Nos. 11 and 17. E. L. WAQNER,J. Chem. Phys. 48,272s (1966). W. J. 0. THOMAS,Tram. Farday Boo. 40, 855 (1963). G. H. KIRKBY, private communication.