Spectroscopy of Hydrothermal Reactions. 8. Kinetics of Aqueous NH4XCN (X = O, S) and OCS by Flow Reactor-Infrared Spectroscopy

Article abstract of DOI:10.1021/ic970723s

The kinetics and pathway of hydrothermolysis of 1 m NH₄SCN to CO₂, NH₃, and H₂S were determined at 543- 573 K and 275 bar by the use of FTIR spectroscopy and a Pt/Ir flow reactor with diamond windows or a 316 stainless steel flow reactor with sapphire windows. The rates of SCN^- loss and CO₂ formation were the same. The reaction is (pseudo) second-order with E_a = 113 ± 11 kg/mol and ln(A, kg/(mol·s)) = 21 ± 2. ΔS? = -84 J/(mol · K), which suggests a bimolecular, rate-determining, initial decomposition step for NH₄SCN. A reaction scheme is proposed in which OCS and a monothiocarbamate species are undetected intermediates. The absence of OCS is explained by the rapid hydrothermolysis rate of OCS to CO₂ and H₂S which was determined by IR spectroscopy at 393-423 K under 275 bar to be E_a = 44 ± 5 kJ/mol and ln(A/s) = 13 ± 1 for OCS. The resulting rate is about 10³ times faster than the hydrothermolysis rate of NH₄SCN at 543 K. The results are compared to the equivalent reaction for NH₄OCN. NH₄OCN reacts about 3 × 10³ times faster than NH₄SCN at 543 K. The trend in the rates is consistent with the charge distribution and the trend in the bond distances, which resulted from ab initio quantum mechanical calculations at the HF/N311G//HF/N31G level in the OCN^- and SCN^- ions and the proposed carbamate and monothiocarbamate intermediates.

Full text of DOI:10.1021/ic970723s

454
Inorg. Chem. 1998, 37, 454-458
Spectroscopy of Hydrothermal Reactions. 8. Kinetics of Aqueous NH4XCN (X ) O, S) and
OCS by Flow Reactor-Infrared Spectroscopy
P. G. Maiella and T. B. Brill*
Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716
ReceiVed June 12, 1997
The kinetics and pathway of hydrothermolysis of 1 m NH4SCN to CO2, NH3, and H2S were determined at 543-
573 K and 275 bar by the use of FTIR spectroscopy and a Pt/Ir flow reactor with diamond windows or a 316
stainless steel flow reactor with sapphire windows. The rates of SCN- loss and CO2 formation were the same.
The reaction is (pseudo) second-order with Ea ) 113 ( 11 kg/mol and ln(A, kg/(mol‚s)) ) 21 ( 2. ∆Sq ) -84
J/(mol‚K), which suggests a bimolecular, rate-determining, initial decomposition step for NH4SCN. A reaction
scheme is proposed in which OCS and a monothiocarbamate species are undetected intermediates. The absence
of OCS is explained by the rapid hydrothermolysis rate of OCS to CO2 and H2S which was determined by IR
spectroscopy at 393-423 K under 275 bar to be Ea ) 44 ( 5 kJ/mol and ln(A/s) ) 13 ( 1 for OCS. The
resulting rate is about 103 times faster than the hydrothermolysis rate of NH4SCN at 543 K. The results are
compared to the equivalent reaction for NH4OCN. NH4OCN reacts about 3 × 103 times faster than NH4SCN at
543 K. The trend in the rates is consistent with the charge distribution and the trend in the bond distances, which
resulted from ab initio quantum mechanical calculations at the HF/N311G//HF/N31G level in the OCN- and
SCN- ions and the proposed carbamate and monothiocarbamate intermediates.
Introduction
destruction efficiencies of smoke and dye compositions,11 and
of DMSO and mercaptans2 are known at SCWO conditions.
Mechanism studies of hydrothermolysis of organosulfur com-
pounds have been reported,12,13 and bond homolysis mechanisms
in sulfur mustard have been explored with ab initio Gaussian
92 calculations in a “solvent” field of various dielectric constants
including those at hydrothermal conditions.14
Although the sulfur-carbon bond is fundamentally important
in numerous areas, investigations of the hydrothermolysis of
organosulfur compounds have been mostly motivated by practi-
cal issues. One of these issues is the destruction of organosulfur
compounds in aqueous waste streams by hydrolysis and
oxidation at high temperatures and pressures. Another issue is
the desulfurization of biomass materials. Two process technolo-
gies that employ the reactivity of H2O at high pressure and
temperature are wet air oxidation,1 where temperatures of 400-
600 K and pressures of 5-200 bar are used, and supercritical
water oxidation (SCWO),2 where temperatures above 675 K
and pressures above 250 bar are used. The critical point of
water is 647 K and 221 bar.
Most previous hydrothermal studies are based on batch
reaction methods with ex situ determination of the products. In
light of the ubiquity of compounds containing the S-C bond
in natural and industrial products, we sought to determine the
kinetics and mechanism of hydrothermolysis of such a bond
by direct spectroscopic measurements. The thiocyanate ion,
SCN-, was chosen for study (as the NH4 salt) for several
+
reasons. First, from the practical point of view the SCN- ion
is present in coke oven gas wastewater3,4 and is considered to
be non-biodegradable.15 Therefore, the kinetics and mechanism
of hydrothermolysis have potentially immediate consequences
as a destruction method. Second, the intense absorbance of
ν3(SCN-) in the mid-IR provides an easy probe of its concen-
tration in an IR flow cell. Third, SCN- is not expected to
The kinetics and mechanisms of hydrolysis of sulfur-
containing chemical warfare agents in the normal liquid range
of H2O have been reviewed.3 The kinetics of removal and/or
the destruction efficiency of organosulfur compounds in the
presence of H2O at high pressure and temperature have been
determined by applying wet air oxidation conditions to steel
mill4,5 and pulp mill6 wastes, pesticides,7 and coal.8-10 The
(7) Randall, T. L. Industrial Waste, Proceedings of the 13th Mid-Atlantic
Conference; Huang, C. P., Ed.; Ann Arbor Science Publishers: Ann
Arbor, MI, 1981; p 501.
(8) Slagle, D.; Shah, Y. T.; Joshi, J. B.; Brainard, A. J. Ind. Eng. Proc.
Res. DeV. 1980, 19, 294.
(9) Lagonik, F. E.; Shah, Y. T.; Albal, R. S.; Joshi, J. B.; Brainard, A. J.
Chem. Eng. Commun. 1981, 11, 201.
* Correspondence author.
(1) Mishra, V. S.; Mahajani, V. V.; Joshi, J. B. Ind. Eng. Chem. Res.
1995, 34, 2.
(2) Tester, J. W.; Holgate, H. R.; Armellini, F. J.; Webley, P. A.; Killilea,
W. R.; Hong, G. T.; Barner, H. E. In Emerging Technologies in
Hazardous Waste Management III; ACS Symposium Series 518,
Tedder, D. W., Pohland, F. G., Eds.; American Chemical Society:
Washington, DC, 1993; p 35.
(10) Joshi, J. B.; Shah, Y. T. Fuel 1981, 60, 612.
(11) Rice, S. F.; LaJeunesse, C. A.; Hanush, R. G.; Aiken, J. D.; Johnston,
S. C. SAND94-8209, Sandia National Laboratories: Livermore, CA,
Jan 1994.
(3) Yang, Y.-C.; Baker, J. A.; Ward, J. R. Chem. ReV. 1992, 92, 1729.
(4) Chowdhury, A. K.; Copa, W. C. Ind. Chem. Eng. 1986, 28, 3.
(5) Copa, W. M.; Gitchel, W. B. Standard Handbook of Hazardous Waste
Treatment and Disposal; Freeman, H. M., Ed.; McGraw-Hill: New
York, 1989; Sec. 8.8.
(6) Schoeffel, E. W.; Seegert, N. Wasser, Luft Betr. 1966, 8, 541; Chem.
Abstr. 1966, 65, 18317.
(12) Abraham, M. A.; Klein, M. T. Fuel Sci. Tech. Int. 1988, 6, 633.
(13) Katritzky, A. R.; Lapucha, A. R.; Greenhill, J. V.; Siskin, M. Energy
Fuels 1990, 4, 562-571. Katritzky, A. R.; Lapucha, A. R.; Luxam,
F. J.; Greenhill, J. V.; Siskin, M. Energy Fuels 1990, 4, 572-577.
Katritzky, A. R.; Murugan, R.; Siskin, M. Energy Fuels 1990, 4, 577.
(14) Politzer, P.; Habibollahzadeh, D. J. Phys. Chem. 1994, 98, 1576.
(15) Horak, O. Chem. Ing. Tech. 1990, 6, 555.
S0020-1669(97)00723-4 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/22/1998

Spectroscopy of Hydrothermal Reactions
Inorganic Chemistry, Vol. 37, No. 3, 1998 455
produce a precipitate or other deleterious products that are
incompatible with a micro flow reactor. Fourth, kinetic data
for hydrothermolysis of the congener salt, NH4OCN, have
recently been determined by the same IR spectroscopic
method,16,17 so that the hydrothermolysis rates of NH4SCN and
NH4OCN can be directly compared for the first time.
It is reasonable that the reactions studied here might be
catalyzed by acid and/or base. By the determination of the
hydrothermolysis kinetics of the salt in water mostly in the
absence of added acid or base, the lower limit of the reaction
rate is revealed, which has practical value in the degradation
process.
Experimental Section
The 90/10 Pt/Ir-diamond flow cell16 and the 316 SS-sapphire cell17
function as the reactor as well as the observation port in the system.
They are transmission IR spectroscopy cells having a path length (30
( 5 µm) that is controlled by the thickness of an Au-foil washer. The
exact path length ((1 µm) was determined by the interference fringe
spacing in the IR spectrum. No detectable change in the path length
was observed at the different conditions used. The heat transfer and
fluid mechanics characteristics of these cells have been determined,16
so that the temperature profiles and flow patterns are known. Plug
flow conditions can be assumed at the higher flow rates (lower degree
of conversion); however, laminar flow is a better description at lower
flow rates.
Figure 1. Mid-IR spectra of 1.00 m NH4SCN at 275 bar and 593 K
at several residence times in the Pt/Ir-diamond window flow cell.
The flow controls for the cells were described previously.17 Briefly,
the flow rate was controlled with an accuracy of (0.01 mL/min by
the output of a pulseless LDC-Analytical, dual-piston, HPLC pump.
The flow rate was monitored continuously by an Intek-Rheotherm
calorimetric flow meter. The temperature was maintained in the (543-
593) ( 1 K range by electrical cartridge heaters mounted concentrically
in the cell block. Active control of these heaters was achieved with
Omega PID control units connected to 0.8 mm K-type thermocouples
mounted in the cell block near the fluid path. The pressure was
maintained at 275 ( 1 bar by an air-actuated bleed-and-lock valve
system. The pressure was activity monitored with an Omega melt
pressure transducer. Software written in Visual Basic was used to
control and record the temperature, pressure, and flow rate at all times.
IR spectra were taken only after the system had stabilized at each set
of conditions.
A 1.00 m (mol/kg) solution of NH4SCN (Aldrich) was prepared by
dissolving the colorless crystals in HPLC-grade H2O that had been
sparged with Ar for 30 min to remove dissolved atmospheric gases.
The hydrothermolysis reaction was characterized in 10 K intervals over
543-573 K using about 15 different flow rates at each temperature.
Flow rates of 1.25-0.05 mL/min yielded residence times in the cell of
2.5-49 s. At each set of conditions, 32 interferograms were summed
at 4 cm-1 resolution using a Nicolet 60 SX FTIR spectrometer
employing a liquid N2-cooled MCT detector. Each spectrum was
ratioed against the spectrum of pure H2O at the same pressure and
temperature to remove as much of the absorbance by H2O as possible.
The spectral features revealed the existence of a single (liquid) phase
at all of the conditions used. Figure 1 illustrates several spectra that
are discussed in this article.
Figure 2. Pseudo-second-order rate plots for the disappearance of
SCN- in 1.00 m NH4SCN at 275 bar (Table 1).
Table 1. Pseudo-Second-Order Rate Constants for
Hydrothermolysis of NH4SCN (CO2 Formation at 275 bar)
T, K
k, kg/(mol‚s)
T, K
k, kg/(mol‚s)
543
553
0.022 ( 0.009
0.027 ( 0.004
563
573
0.046 ( 0.004
0.066 ( 0.006
concentrations by using the calibration data and were corrected for the
change in density of H2O [0.795 g/cm3 (543 K) to 0.747 g/cm3 (573
K)].
Conversion of the spectral absorbance areas into concentrations
required acquisition of calibration data based on the Lambert-Beer
law. ν3(CO2) for the CO2-H2O system has been characterized for this
purpose in the infrared region,17 but adjustment of the absorbance values
of CO2 to match the required carbon balance at the end of the reaction
corrects for small variations in the volume of the different cells and
differences in the cell alignment.18 The spectral band areas were then
determined by using curve fitting software (Peakfit, Jandel Scientific)
with a four-parameter Voigt function. These areas were converted to
The reaction was determined to have a second-order rate law on the
basis of the linear plot of 1/[SCN-] vs time (Figure 2). The
concentration of SCN- was determined from the IR absorbance and
also by the carbon mass balance based on the rate of formation of CO2.
Both gave the same rate within the uncertainty of the experiment. The
CO2 data were used because we have experience with the CO2 line
shape and absorptivity in H2O up to 350 °C.17,18 Table 1 contains the
resulting rate constants.
+
The reaction was determined to be first order with respect to NH4
and SCN- by adding known concentrations of NH4NO3 to vary the
NH4+ concentration and KSCN to vary the SCN- concentration. These
solutions were run in triplicate at 553 K and 275 bar for 13 residence
times. This temperature is below the decomposition temperature of
aqueous NH4NO3 in our flow cell.16 Table 2 contains the observed
(16) Schoppelrei, J. W.; Kieke, M. L.; Wang, X.; Klein, M. T.; Brill, T. B.
J. Phys. Chem. 1996, 100, 14343.
(17) Kieke, M. L.; Schoppelrei, J. W.; Brill, T. B. J. Phys. Chem. 1996,
100, 7455.
(18) Maiella, P. G.; Brill, T. B. Unpublished results.

456 Inorganic Chemistry, Vol. 37, No. 3, 1998
Maiella and Brill
Table 3. Arrhenius Parameters in H2O at 275 bar
Ea, kJ/
+
Table 2. Effect on the Rate of CO2 Formation of Excess NH4 and
SCN- in 1.0 m NH4SCN (T ) 553 K, P ) 275 bar)
ln(A, kg/
(mol‚s))
concn. kg/mol
reacn
NH4OCN (eq 2) Pt/Ir-diamond
NH4OCN (eq 2) 316 SS-sapphire 473-573 59
cell
T, K
mol
NH4SCN
NH4NO3
KSCN
kobsd, kg/(mol‚s)
kobsd/[X]
473-573 67.5 ( 7 18.6 ( 0.3
17.1
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0
0
0
0
0
0.5
0.75
1.0
0.027 ( 0.004
0.038 ( 0.005
0.046 ( 0.007
0.050 ( 0.005
0.043 ( 0.009
0.049 ( 0.003
0.055 ( 0.008
0.027
0.025
0.026
0.025
0.029
0.028
0.028
NH4SCN (eq 3) 316 SS-sapphire 543-573 113 ( 11 21 ( 2
0.5
0.75
1.0
0
0
0
OCS (eq 7)
316 SS-sapphire 393-423 44 ( 5
13 ( 1a
a Units are s-1
.
were acquired by modeling the hydrothermolysis of 1.05 m urea
at 473-573 K in which the hydrothermolysis of NH4OCN is
one of the observable steps. The most consistent rate constants
and Arrhenius parameters were extracted from the species
profiles of eqs 1 and 2 by using the appropriate, iteratively-
rate constants calculated from the appropriate rate expression19 when
excess NH4+ and excess SCN- have been added. The kinetic analysis
was limited to residence times of 10 s or less so that plug flow
conditions could be assumed. Weighted least-squares regressions were
used to determine all of the rate constants and Arrhenius parameters.
The statistical weight ωi used was 1/σ2, where σ is the standard
deviation. Where applicable, ωi was approximated as k2ωi.20
The products of hydrothermolysis of NH4SCN were also identified
after heating the solution in a sealed 12 cm3 316 SS tube in a fluidized
sand bath. By the trapping of the gases released when the tube was
opened, the existence of H2S, which was not detected in the IR spectrum
due to its low molar absorptivity, was confirmed by its mass spectrum
(m/z ) 32, 33, 34).
The kinetics of hydrothermolysis of carbonyl sulfide were determined
on an aqueous solution that had been saturated with OCS (Matheson)
by bubbling and stirring in an autoclave for approximately 175 min at
296 K. The solutions had a concentration of approximately 0.25 m
based on total conversion to CO2. This solution was pumped in the
closed flow reactor as described above, and the rate of conversion to
CO2 was measured at 393-423 K under 275 bar.
We had hoped to include the hydrothermolysis of NH4SeCN in this
study for comparison with NH4OCN and NH4SCN. NH4SeCN was
synthesized,21 but the plan was abandoned when it was found by batch
reaction that Se deposited on the tube walls. Such an occurrence would
plug the flow reactor. Although the mechanism of hydrothermolysis
of NH4SeCN was not determined, the formation of solid Se may result
from the H2Se h H2 + Se equilibrium.
Ab initio quantum mechanical calculations were performed with the
General Atomic and Molecular Electronic Structure System (GAMESS).22
The energy gradient method was used to optimize the geometry of each
species. Urea, carbamate, and cyanate were optimized by using the
HF/311G basis set. The HF/31G basis set was used for thiourea,
monothiocarbamate, and thiocyanate. Vibrational frequencies were
calculated using second derivatives at the forementioned levels to
confirm the stationary point structures. Second-order Møller-Plesset
perturbation calculations (MP2)23,24 at the forementioned levels were
executed to account for the effects of electron correlation in the total
energies.
(NH2)2CO f NH4+ + OCN-
(1)
(2)
NH4+ + OCN- (+H2O) f CO2 + 2NH3
solved, differential rate expressions.16,17 Table 3 gives the
Arrhenius parameters for the pseudo-second-order eq 2 deter-
mined with cells constructed from two different materials. The
differences in the NH4OCN data can be attributed to error in
the assumed cell volume and/or, possibly, to a small contribution
from the difference in the materials of construction.
Ammonium Thiocyanate Spectra and Kinetics. Unlike
NH4OCN, the hydrothermolysis reaction of 1.00 m NH4SCN
could be followed directly in both the 316 SS-sapphire cell and
the Pt/Ir-diamond cell at a constant pressure of 275 bar as a
function of temperature and flow rate. This was possible
because NH4SCN reacted more slowly than NH4OCN. The
temperature range used for kinetics was 543-573 K, although
data were collected up to 593 K. Figure 1 shows spectra
selected from experiments with the Pt/Ir-diamond cell. Absor-
bances for the SCN- and NH4+ reactant ions are clearly apparent
at short residence times in which the reaction is at an early stage.
They diminish in intensity at longer residence times and are
replaced by the absorbance of aqueous CO2. Apart from the
appearance of a broad, weak absorbance centered at about 1100
cm-1 at the highest temperature used, no other products are
apparent. Although the exact origin of this latter absorbance
is unknown at this time, it appears in the intensely IR-active
S-O stretching region. This suggests that oxidation of the H2S
product occurs. The amount of oxidation is apparently small
because the intensity is relatively low. Extensive oxidation of
sulfur is known to occur when an S-H2O mixture is held at
638 K in the batch mode for longer times.25
Studies in which the concentrations of NH4+ and SCN- were
independently varied showed that the rate of hydrothermolysis
Results and Discussion
of NH4SCN was essentially first-order in NH4 and in SCN-
+
Ammonium Cyanate Data. The kinetics of hydrothermoly-
sis of NH4OCN have been determined by IR spectroscopy in
the same Pt/Ir-diamond and 316 SS-sapphire cells as were used
to investigate NH4SCN.16,17 It was not feasible, however, to
determine the hydrothermolysis kinetics of NH4OCN directly
by simply dissolving the salt in H2O, because it decomposed
rapidly in the temperature range of interest. Instead, the kinetics
(Table 2). Thus, the rate should be second-order in NH4SCN,
which is consistent with eq 3 and was validated by plotting
NH4+ + SCN- + 2H2O f CO2 + 2NH3 + H2S (3)
1/(C0 - [CO2]) vs the residence time as shown in Figure 2.
Because of the probable role of H2O in the reaction, these rates
(Table 1) are best described as pseudo-second-order. A plot of
the rate constants at 543-573 K yields the Arrhenius plot shown
in Figure 3 and Arrhenius parameters shown in Table 3. Also
shown in Figure 3 is the equivalent Arrhenius plot for NH4-
(19) Connors, K. A. Chemical Kinetics; VCH Publ. Inc.: New York, 1990.
(20) Cvetanovic, R. J.; Singleton, D. L. Int. J. Chem. Kinet. 1977, 9, 481.
(21) Geisler, K.; Nobst, E.; Mu¨ller, C.; Bulka, E. Z. Chem. 1982, 22, 113.
(22) Schmidt, M. W.; Baldridge, K. K.; Boarz, J. A.; Elbert, S. T.; Gordon,
M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su,
S. J.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem.
1993, 14, 1347.
(25) Sorokin, V. I.; Orlov, R. Yu.; Dadze, T. P. Proceedings of the First
International Conference on SolVo-thermal Reactions; Shikoku Na-
tional Industrial Research Institute: Takamatsu, Japan, Dec. 1994;
paper S-5.
(23) Binkley, J. S.; Pople, J. A.; Hehre, W. J. J. Am. Chem. Soc. 1975, 97,
229.
(24) Pople, J. A.; Hehre, W. J. Int. J. Quantum Chem. Symp. 1976, 10, 1.

Spectroscopy of Hydrothermal Reactions
Inorganic Chemistry, Vol. 37, No. 3, 1998 457
Figure 3. Arrhenius plots for decomposition of aqueous NH4SCN,
NH4OCN, and OCS at 275 bar.
OCN for comparison with NH4SCN. It is obvious that the rate
of hydrolysis of NH4SCN is slower than that of NH4OCN under
the same conditions.
Figure 4. Mid-IR spectra of approximately 0.25 m OCS during
hydrothermolysis at 493 K and 275 bar at several residence times in
the 316 SS flow reactor.
An indication of the molecularity of these reactions is
provided by ∆Sq. At 543 K in the 316 SS-sapphire cell, ∆Sq
) -113 J/(mol‚K) for NH4OCN and -84 J/(mol‚K) for NH4-
SCN. These values compare well to the approximate range for
bimolecular reactions (∼ -63 to -125 J/(mol‚K)),19 which is
consistent with the pseudo-second-order rate expression deter-
mined for eqs 2 and 3. The more negative value of ∆Sq for
NH4OCN is consistent with the greater charge compactness of
the OCN- ion compared to SCN-, which enhances the elec-
trostriction of the surrounding H2O field.
Table 4. Pseudo-First-Order Rate Constants for Hydrothermolysis
of OCS (CO2 Formation at 275 bar)
T, K
k, (s-1
)
T, K
k, (s-1
)
393
403
0.42 ( 0.03
0.70 ( 0.10
413
423
0.80 ( 0.10
1.10 ( 0.12
absent. However, unlike OCN- which should be detected if it
is present,16,17 the absence of OCS was understood after its
kinetics of hydrothermolysis were determined as described next.
Reaction Pathways. Based on the spectra in Figure 1, eqs
4 and 5 are one potential pathway of hydrothermolysis of NH4-
Figure 4 shows IR spectra of the C-O stretching region of
approximately 0.25 molal OCS during the conversion to CO2
by eq 7 at 393 K under 275 bar. Several of many residence
times recorded are shown. As with the NH4SCN solution above,
no H2S could be detected, which is consistent with its low IR
absorptivity. Conversion of the CO2 areas into concentrations
provides the pseudo-first-order rate constants for eq 7 at 393-
423 K under 275 bar. These rate constants are given in Table
4 and provide the Arrhenius plot shown in Figure 3. The
comparatively low activation energy of 44 ( 5 kJ/mol (Table
3) makes the reaction facile despite the fact that the A factor
suggests at least bimolecular character (∆Sq ) -150 J/(mol‚K)).
Bimolecularity is not inconsistent with a pseudo-first-order rate
expression for eq 7. Extrapolation of the rate of hydrother-
molysis of OCS to 543 K, where the slowest rate of decomposi-
tion of NH4SCN was measured, indicates that OCS decomposes
about 103 times faster than NH4SCN. Consequently, the OCS
concentration will be below the limit of detection in our cell.
As a result of these findings, the net hydrothermolysis reaction
for NH4SCN (eq 3), which parallels eq 2 for NH4OCN, is
proposed to be the result of eqs 6 and 7.
NH4+ + SCN- + H2O f NH4+ + OCN- + H2S (4)
NH4+ + OCN- + H2O f CO2 + 2NH3
(5)
SCN giving eq 3. This scheme was firmly rejected on the basis
of the absence of the intensely active stretching mode of OCN-
at 2165 cm-1. This mode is readily detectable in the 543 K
temperature range when OCN- is present at greater than 0.01
m.16,17 On the other hand, the presence of H2S was apparent
by its pungent odor and was confirmed by mass spectrometry
(Vide supra). Quantitation of H2S by mass spectrometry was
not satisfactory probably for the reasons that a two-phase system
exists after depressurization and the fact that some of the H2S
is oxidized (Figure 1). H2S(aq) was expected to be observed
in the IR spectrum by ν(SH) at about 2680 cm-1; however, in
contrast to the O-H stretch of H2O, the S-H stretch of H2S
has low intensity, in part because hydrogen bonding by H2S is
weak. No H2S could be detected by IR spectroscopy even in
an H2O standard solution that had been saturated with H2S.
A more justifiable pathway of hydrothermolysis of NH4SCN
is given by eqs 6 and 7. At first sight this scheme appears to
The rate differences found are helpful for hypothesizing
further details about the pathway of eqs 2 and 3. Figure 5 shows
the proposed scheme. For comparison, the rate constants at
543 K are shown as derived from the Arrhenius data in Table
3. Hydrothermolysis of NH4SCN and NH4OCN plausibly
begins with association of H2O and the electropositive carbon
atom of the ion pair. Incidently, ion-pair formation by salts is
increasingly favored in water at higher temperatures because
the bulk dielectric constant of water decreases with increasing
NH4+ + SCN- + H2O f 2NH3 + OCS
OCS + H2O f CO2 + H2S
(6)
(7)
be no more satisfactory than eqs 4 and 5 because the intensely
IR-active ν3(OCS) at about 2100 cm-1 is, like OCN- in eq 4,

458 Inorganic Chemistry, Vol. 37, No. 3, 1998
Maiella and Brill
KOH solution was added to 1.00 m NH4SCN to raise the pH to
11.25 (at 295 K). The rate constant of hydrothermolysis of NH4-
SCN at 543 K was observed to be 2.2 ( 0.2 x 10-3 kg/(mol.s),
which is one-tenth the value for the 1.00 m NH4SCN solution
(Table 1). No reaction at all was detected at 543 K when the
pH was raised to 14. To conclude discussion of Figure 5, the
absence of OCS in the IR spectrum during the hydrothermolysis
of NH4SCN was explained by the kinetics of hydrothermolysis
of OCS described above. Its rate constant at 543 K is about
103 times faster than the hydrothermolysis rate of NH4SCN at
this temperature.
The comments just made suggested reasons that elimination
of the C-N bond during hydrothermolysis of X-C-N linkages
is impeded when X ) S compared to X ) O. This pattern
also exists in compounds related to NH4OCN and NH4SCN,
such as their compositional isomers urea, (NH2)2CO, and
thiourea, (NH2)2CS. Each has two C-N bonds. The rear-
rangement of these compounds by eqs 828 and 929 is well-known.
(NH2)2CO h NH4OCN
(8)
(NH2)2CS h NH4SCN
(9)
Figure 5. Proposed hydrothermolysis pathway for aqueous NH4SCN
and NH4OCN showing charges and bond parameters (Å) from ab initio
quantum mechanical calculations and rate comparisons at 543 K under
275 bar.
The first-order kinetics of the forward reaction in eq 8 in H2O
have been determined at 333-363 K30 and at 473-573 K.16,17
Kinetics of the forward reaction of eq 9 in H2O are known only
at 363-403 K.31 Thus, the rates of these reactions are most
safely compared at 363 K, and the effect of X ) S or O on
scission of the first of two XC-N bonds is thereby indicated.
For thiourea31 k ) 1.5 × 1014 exp (-18.06/T) s-1. For urea30
k ) 3.8 × 1014 exp (-16.37/T) s-1. The rate of conversion of
thiourea is, therefore, about 250 times slower than urea at 363
K implying, as before, that S impedes the scission of the C-N
bonds in these compounds more than O. In keeping with the
kinetics, the average C-N bond distance in (NH2)2CS is shorter
[1.33 Å (X-ray32 ), 1.37 Å (calcd)] than that in (NH2)2CO [1.34
Å (X-ray33 ), 1.39 Å (calcd)]. The calculated values were
obtained by the ab initio quantum mechanical calculations
described above.
In summary, the first spectroscopic measurements were made
on the decomposition of compounds containing the XCN linkage
(X ) O, S) in the neutral hydrothermal medium. Kinetic
constants reveal that the rate of the reaction is slower when X
) S. While not directly successful in revealing the intermedi-
ates, the IR spectra provide the basis for rationally sorting out
the pathways. The coupling of in situ spectral data and ab initio
quantum mechanical calculations provides reasonable insight
into the mechanism under demanding experimental conditions.
temperature.26 Many possible ways now exist for migration of
hydrogen as a C-O bond forms. The rate of reaction is,
however, expected to be slower for SCN- because a Mulliken
population analysis indicates that less positive charge resides
on the carbon atom in SCN- (+0.003) than OCN- (+0.159).
Hence, the initial H2O‚‚‚C interaction is weaker in SCN-.
Reasonable resulting intermediates of hydrolysis of NH4OCN
and NH4SCN are the carbamic and the monothiocarbamic acids
or their anions, respectively, along with NH3 (Figure 5).
However, the steady-state concentrations of the carbamate and
monothiocarbamate species are expected to be below the limit
of detection at the temperature of this study. This expectation
is based on the fact that, although the carbamate ion can be
detected at as high as 450 K in the Raman spectrum of aqueous
(NH4)2CO3,27 it is not detected at this temperature by IR
spectroscopy.17
The rate of conversion of the monothiocarbamate species to
OCS and NH3 in Figure 5 is expected to be slower than the
carbamate species. This is a consequence of the ab initio
quantum mechanical calculations on NH2C(O)OH and NH2C-
(S)OH that reveal the C-N bond to be shorter in the latter
compound. The same trend is also found in calculations on
the anions (1.42 and 1.46 Å, respectively). We emphasize,
however, that the rate-controlling step of hydrothermolysis of
NH4SCN is proposed to be the initial pseudo-second-order step
but suggest that the second step in Figure 5, which includes
breaking of the C-N bond, may also be slower when S is bound
to C than when O is bound to C.
Acknowledgment. We are grateful to the U. S. Army
Research office for support of this work on Grant DAAL03-
92-G-0174. We thank Steven D. Bennett for help with the
computational methods.
IC970723S
+
(28) Wo¨hler, F. Pogg. Ann. 1828, 12, 253.
(29) Reynolds, J. E. J. Chem Soc. 1869, 22, 1.
(30) Shaw, W. H. R.; Bordeaux, J. J. J. Am. Chem. Soc. 1955, 77, 4729.
(31) Shaw, W. H. R.; Walker, D. G. J. Am. Chem. Soc. 1956, 78, 5769.
(32) Kutoglu, A.; Scheringer, C.; Meyer, H.; Schweig, A. Acta Crystallogr.
1982, B38, 2626.
Because NH4 acts as a proton donor in the initial step of
Figure 5, an increase in the pH should decrease the rate. A
(26) Uematsu, M.; Franck, E. U. Phys. Chem. Ref. Data 1980, 9, 1291.
(27) Schoppelrei, J. W.; Kieke, M. L.; Brill, T. B. J. Phys. Chem. 1996,
100, 7463.
(33) Mullen, D.; Hellner, E. Acta Crystallogr. 1978, B34, 1624.