Experimental and Theoretical Investigation of Homogeneous Gaseous

Article pubs.acs.org/JPCA

Experimental and Theoretical Investigation of Homogeneous Gaseous Reaction of CO2(g) + nH2O(g) + nNH3(g) → Products (n = 1, 2) Zhuangjie Li* and Baoquan Zhang Department of Chemistry and Biochemistry, California State University Fullerton, Fullerton, California 92834, United States S Supporting Information *

ABSTRACT: Decreasing CO2 emissions into the atmosphere is key for reducing global warming. To facilitate the CO2 emission reduction efforts, our laboratory conducted experimental and theoretical investigations of the homogeneous gaseous reaction of CO2(g) + nH2O(g) + nNH3(g) → (NH4)HCO3(s)/(NH4)2CO3(s) (n = 1 and 2) using Fourier transform infrared attenuated total reflectance (FTIR-ATR) spectroscopy and ab initio molecular orbital theory. Our FTIRATR experimental results indicate that (NH4)2CO3(s) and (NH4)HCO3(s) are formed as aerosol particulate matter when carbon dioxide reacts with ammonia and water in the gaseous phase at room temperature. Ab initio study of this chemical system suggested that the reaction may proceed through formation of NH3·H2O(g), NH3·CO2(g), and CO2·H2O(g) complexes. Subsequent complexes, NH3·H2O·CO2 and (NH3)2·H2O·CO2, can be formed by adding gaseous reactants to the NH3·H2O(g), NH3·CO2(g), and CO2·H2O(g) complexes, respectively. The NH3·H2O·CO2 and (NH3)2·H2O·CO2 complexes can then be rearranged to produce (NH4)HCO3 and (NH4)2CO3 as final products via a transition state, and the NH3 molecule acts as a medium accepting and donating hydrogen atoms in the rearrangement process. Our computational results also reveal that the presence of an additional water molecule can reduce the activation energy of the rearrangement process. The high activation energy predicted in the present work suggests that the reaction is kinetically not favored, and our experimental observation of (NH4)HCO3(s) and (NH4)2CO3(s) may be attributed to the high concentrations of reactants increasing the reaction rate of the title reactions in the reactor.

■

atmosphere and point sources.5−17 Previous CO2 removal methods have focused on (1) absorption of CO2 into aqueous solutions, including NH3 solution,9,11 K2CO3 solution,10 aqueous blend of 2-amino-2-methyl-propanol and monoethanolamine,16 and aqueous red mud;6 and (2) absorption of CO2 on solid sorbents, such as CaO,12 liquid-impregnated clay,7 calciumbased sorbents,14 and molecular sieve.8 It has also been proposed to remove atmospheric CO2 by increasing oceanic uptake of CO2 molecules through making ocean water more alkaline by removing HCl from the ocean water, 13 and by ocean sequestration of crop residue carbon through burying the carbon into the deep ocean.5 These techniques for CO2 removal are primarily based on physical adsorption or chemical conversion of CO2 in solution or on solid surfaces, and many of these methods only collect the CO2 without its significant utilization. Sustainable removal of CO2, which would require both conversion and utilization of the molecule, could reduce global warming and benefit the economy.15 This purpose renders a desirable process that could remove CO2 by transforming this molecule into industrially useful chemicals via gaseous phase

INTRODUCTION Carbon dioxide, CO2, has been known to be one of the main atmospheric species affecting climate change. Its primary contribution to global warming arises from its ability to absorb infrared radiation emitted from the Earth and retain the infrared energy in the atmosphere. Previous studies have suggested that anthropogenic emission of CO2 has significantly increased over the past century, contributing to approximately 2/3 of the present day CO2-induced warming.1,2 A recent climate model study has shown that the climate change caused by the increase of atmospheric CO2 abundance could lead to sea level rise and irreversible changes in global precipitation patterns, including those adversely affecting dry-season rainfall in the southwest of United States.3 Further model studies of global temperature response to atmospheric CO2 increase predicted that the global temperature would rise 1.0−2.1 °C per trillion tons (Tt C) of carbon emitted, and that total carbon emission must be restricted to about 0.8 Tt C in order to control the warming temperature within 2 °C relative to preindustrial temperature.4 Therefore, it is important to reduce CO2 emission to minimize the stress on our environmental habitat and to avoid the catastrophic consequences that would likely result from global warming. There has been great interest in developing approaches and techniques for sequestration and removal of CO2 from the © 2012 American Chemical Society

Received: April 20, 2012 Revised: August 14, 2012 Published: August 20, 2012 8989

dx.doi.org/10.1021/jp303848h | J. Phys. Chem. A 2012, 116, 8989−9000

The Journal of Physical Chemistry A

Article

material, and can be used to collect the infrared spectra of small amounts (99.9%) was used as the water vapor source.

reactions that require no energy and do not lead to secondary environmental pollution. Gaseous phase reaction of CO2 with water and ammonia has been reported to produce solid products, (NH4)HCO3(s) and (NH4)2CO3(s):9,17 CO2 (g) + H 2O(g) + NH3(g) → (NH4)HCO3(s)

(1)

CO2 (g) + H 2O(g) + 2NH3(g) → (NH4)2 CO3(s)

(2)

However, the products of these gaseous phase reactions have not been analyzed using infrared spectroscopy, and the reaction mechanism of these chemical processes is not clearly understood. In this paper we report our experimental and theoretical investigations of reactions 1 and 2 using Fourier transform infrared attenuated total reflectance (FTIR-ATR) spectroscopy and ab initio molecular orbital theory. Our experimental work demonstrates that these reactions do occur at room temperature with production of (NH4)HCO3(s) and (NH4)2CO3(s) in the form of aerosols, and our theoretical work explores the potential energy surfaces of reactions 1 and 2, with and without the presence of an additional water molecule. We will then comment, based on the findings of the present work, on the mechanism of these reactions and the implication of these gaseous phase chemical processes to the global and climate changes.

■

EXPERIMENTAL AND THEORETICAL DETAILS The apparatus used in the present work is shown in Figure 1. The reactor was made of a 6 in. long Pyrex tube with 1.0 in. i.d. where

■

RESULTS AND DISCUSSION a. Experimental Observations. Figure 2 shows snapshots of the experiment for the reaction of CO2(g) with H2O(g) plus NH3(g) at room temperature. It can be seen from Figure 2a that when ∼15 Torr of water vapor was mixed with ∼447 Torr of carbon dioxide in the reactor, the reactor appearance did not change, indicating that there was little interaction between the CO2(g) and H2O(g). When about 18 Torr of NH3(g) was added to the reactor containing the mixture of CO2(g) and H2O(g), white aerosol particles were observed to form in both the gas phase and on the wall of the reactor (t = 0−6200 ms snapshot). Figure 2b, which displays the total pressure (in Torr) inside the reactor as a function of time, indicates that reactor pressure decreased at about 4 s as a result of consumption of the reactants due to reactions 1 and 2. Additional experiments were performed under different gaseous reactant pressures (Table 1) to confirm the formation of the aerosol products. The aerosol particles were also observed to form when three reactants were introduced into the reactor in different orders, as well as when the vapor from a 29.9% ammonia solution was added to the reactor containing ∼200 Torr of CO2(g). Figure 3 shows the FTIR-ATR spectrum of the solid products collected on the wall of the reactor along with that of commercially available (NH4)2CO3(s) and (NH4)HCO3(s). The frequencies of major peaks are listed in Table 2. The FTIRATR spectra of (NH4)HCO3(s) and (NH4)2CO3(s) are similar to each other, which is consistent with these compounds’ FTIR absorption spectra.19 However, modest differences still occur between the exact frequencies of the two compounds. The peaks in the frequency range of 600−1800 cm−1 are assigned to the bending and stretching vibrational motions of the carbonate

Figure 1. Experimental schematic for the reaction of CO2(g) + H2O(g) + NH3(g).

the reaction of CO2(g) + NH3(g) + H2O(g) took place. The tube was first cleaned using water, rinsed three times with 5 mL acetone each, and finally pumped down to less than 10 mTorr prior to introducing the reactants. The baseline pressure was essentially free of water and CO2 since the tube only had acetone residue remaining after being cleaned. During the experiments, the reactant gases were individually added to the reactor in several different orders to verify the observations. The pressure of each reactant in the reactor was monitored using a Baratron capacitance manometer (MKS 627B13TAC1B 1000 Torr). Solid product was collected by scraping the solid substance coated on the wall of the reactor with a spatula, and the solid sample was analyzed using an FTIR spectrometer (Thermo Nicolet 4700) equipped with a single bounce ATR assembly (SMART PERFORMER). The ATR assembly is a standard accessory device using a ZnSe crystal as the total reflectance 8990



Article

Table 1. Gaseous Reactant Pressure (in Torr) of the Experimental Runs for the Reaction of NH3(g) + H2O(g) + CO2(g)a run number

PH2O

PCO2

PNH3b

1 2 3 4

5.8 (0) 5.0 (0) 10.1 (0) 15.4 (0)

209 (1.1) 102 (0.8) 290 (4.2) 447 (0.7)

203 (1.3) 255 (1.3) 114 (4.3) 40 (3.7)

a

The values in parentheses are the relative time (in minutes) at which the reactant gases were introduced into the reactor. bThe ammonia pressure was estimated from the total pressure change of the reactor after a pulse of the ammonia was injected into the reactor.

Figure 3. The FTIR-ATR spectrum of the solid products (blue) collected from the reaction of CO2(g) + H2O(g) + NH3(g) along with the FTIR-ATR spectrum of (NH4)HCO3 (green) and (NH4)2CO3 (red) reference compounds. The spectral resolution was 4 cm−1.

Table 2. Infrared Vibrational Frequencies (in cm−1) of Solid Products in Comparison to the Vibrational Frequencies of NH4HCO3, and (NH4)2CO3 in the Frequency Range of 600− 3500 cm−1 compounds NH4HCO3 (NH4)2CO3 solid products

frequency (cm−1) 663, 687, 831, 949, 1024, 1119, 1259, 1329, 1438, 1477, 1495, 1595, 2837, 3051, 3170 665, 685, 830, 949, 1022, 1128, 1333, 1398, 1446, 1537, 1626, 2848, 3051, 3170, 3454 667, 686, 827, 955, 1026, 1119, 1348, 1394, 1450, 1531, 1622, 2862, 3047, 3178, 3464

solid product FTIR-ATR spectrum also has several broad bands at 2700−3300 cm−1 resulting from vibrational motions associated with the NH3 group. Note that a moderately broad peak occurs at 3464 cm−1 in our solid product FTIR-ATR spectrum, which is slightly greater than the one belonging to ammonium carbonate, which occurs at 3454 cm−1. This moderately broad peak suggests the presence of ammonium carbamate (NH2COONH4),19 which has also been identified by Li et al. (2003) using NMR spectroscopy.17 While the percentages of ammonia carbonate, ammonia bicarbonate, and ammonium carbamate in the solid product remain to be determined, our experimental data indicate that both reactions 1 and 2 take place when CO2(g), NH3(g), and H2O(g) are mixed in the reactor.

Figure 2. Snapshots of a video movie recording the reaction of CO2(g) + H2O(g) + NH3(g) at room temperature (a) and the total pressure (in Torr) change inside the reactor as a function of time (b).

group of both ammonium bicarbonate and carbonate on the basis of our theoretical calculation predictions (Table 5 of the Supporting Information), which is also consistent with most previous general assignments of these frequencies.19 Our spectral analysis suggests that the solid products include (NH4)2CO3(s) and (NH4)HCO3(s), as indicated by absorption peaks at 667, 686, 827, 955, 1026, 1119, 1348, 1450, 1531, and 1622 cm−1. The 8991



Article

Figure 4. continued

8992



Article

Figure 4. Optimized structures for species involved in the reaction of CO2(g) + nH2O(g) + nNH3(g) → (NH4)HCO3/(NH4)2CO3(s) (n = 1 and 2). The bond lengths are in Ǻ and bond angles and dihedral angles expressed by D (numbered atoms a,b,c,d) are in degrees. The numbers in parentheses are experimental values.

b. Theoretical Results: Without the Presence of One Additional Water Molecule. It is unclear how the reaction involving CO2(g), NH3(g), and H2O(g) proceeded in gaseous phase to result in our experimental observations. Furthermore, the formation of the solid products could have resulted from heterogeneous processes, and it is possible that the investigated reaction was heterogeneous in nature. Since these issues cannot be addressed experimentally in the present investigation, theoretical examination of this chemical system may help shed light on the reaction mechanism for the experimental observations.

Possible reaction mechanisms for gaseous phase reactions 1 and 2 can be postulated as follows: Reaction 1 Pathway I NH3 + H 2O ⇌ NH3·H 2O NH3·H 2O + CO2 ⇌ NH3·H 2O·CO2 NH3·H 2O·CO2 ⇌ TS1 → NH4HCO3 8993



Article

Table 3. Calculated Total Energy (in Hartree) and Zero Point Energy (in kcal mol−1) for the Species Involved in the NH3(g), H2O(g), and CO2(g) Chemical System

NH3 H2O CO2 NH3·H2O H2O·CO2 NH3·CO2 NH3·H2O·CO2 TS1 NH4HCO3 (NH3·H2O·CO2)·H2O TS1·H2O (NH4HCO3)·H2O (NH3)2·H2O (NH3)2·CO2 (NH3)2·H2O·CO2 TS2 (NH4)2CO3 ((NH3)2·H2O·CO2)·H2O TS2·H2O ((NH4)2CO3)·H2O

MP2

MP4a

6-311G(d,p)

6-311++G(d,p)

CBS-QB3b (enthalpy)

zero point energyc

−56.40879 −76.26397 −188.19913 −132.68587 −264.46868 −244.61380 −320.89396 −320.84610 −320.88156 −397.17896 −397.13420 −397.16837 −189.10047 −301.03305 −377.31507 −377.27694 −377.31352 −453.60205 −453.56192 −453.59787

−56.43451 −76.28721 −188.23251 −132.73318 −264.52523 −244.67258 −320.97386 −320.92384 −320.95662 −397.27581 −397.22961 −397.26187 −189.17372 −301.11586 −377.41822 −377.37624 −377.41099 −453.72278 −453.68095 −453.71535

−56.45575 −76.33314 −188.36848 −132.79639 −264.70489 −244.82822 −321.17086 −321.13112 −321.16258 −397.51489 −397.48189 −397.51123 −189.25506 −301.28991 −377.63347 −377.60412 −377.63515 −453.97875 −453.95149 −453.98258

21.9 13.7 37.9 7.3 21.9 30.0 46.3 49.1 46.7 63.0 65.5 63.0 70.0 53.8 70.6 73.3 61.7 87.0 86.7 89.4

a

Single point calculation using the geometry optimized at the MP2/6-311G(d,p) level of theory. bCBS-QB3 calculations started with the optimized structures and frequencies computed at the MP2/6-311G(d,p) level of theory. cZero point energy was calculated at the MP2/6-311G(d,p) level of theory.

Pathway II

Pathway III

CO2 + H 2O ⇌ CO2 ·H 2O

CO2 + H 2O ⇌ CO2 ·H 2O

CO2 ·H 2O + NH3 ⇌ NH3·H 2O·CO2


NH3·H 2O·CO2 ⇌ TS1 → NH4HCO3

NH3·H 2O·CO2 + NH3 ⇌ (NH3)2 ·H 2O·CO2 (NH3)2 ·H 2O·CO2 ⇌ TS2 → (NH4)2 CO3

Pathway III CO2 + NH3 ⇌ NH3·CO2

Pathway IV

NH3·CO2 + H 2O ⇌ NH3·H 2O·CO2

CO2 + NH3 ⇌ NH3·CO2

NH3·H 2O·CO2 ⇌ TS1 → NH4HCO3

NH3·CO2 + NH3 ⇌ (NH3)2 ·CO2 (NH3)2 ·CO2 + H 2O ⇌ (NH3)2 ·H 2O·CO2

Reaction 2

(NH3)2 ·H 2O·CO2 ⇌ TS2 → (NH4)2 CO3

Pathway I

A major difference between the postulated pathways for reactions 1 and 2 occurs in the formation of the NH3·H2O complex versus the formation of the CO2·H2O and NH3·CO2 complexes in the initial step of the reaction. In the case of reaction 2, another difference lies in the addition of either CO2 or NH3 onto the NH3·H2O, CO2·H2O and NH3·CO2 complexes in the second step to form subsequent complexes. To evaluate the most energetically favorable pathway leading to the formation of NH4HCO3 and (NH4)2CO3, ab initio calculations were carried out for each of the species involved in the above pathways. Our calculations focused on the structural optimization of the complexes and the energy associated with these species, which are essential for mapping the potential energy surfaces for the postulated reaction pathways. The calculated structures of these species are shown in Figure 4. The absolute energy associated with the optimized structure of each species is given in Table 3, and the relative energy with

NH3 + H 2O ⇌ NH3·H 2O NH3·H 2O + CO2 ⇌ NH3·H 2O·CO2 NH3·H 2O·CO2 + NH3 ⇌ (NH3)2 ·H 2O·CO2 (NH3)2 ·H 2O·CO2 ⇌ TS2 → (NH4)2 CO3

Pathway II NH3 + H 2O ⇌ NH3·H 2O NH3·H 2O + NH3 ⇌ (NH3)2 ·H 2O (NH3)2 ·H 2O + CO2 ⇌ (NH3)2 ·H 2O·CO2 (NH3)2 ·H 2O·CO2 ⇌ TS2 → (NH4)2 CO3 8994



Article

Table 4. Computed Relative Energy (in kcal mol−1) for Reactions of CO2(g) with NH3(g) and H2O(g) MP2/6-311G(d,p)+ΔZPE

PMP4/6-311++G(d,p)+ΔZPE

CBS-QB3+ΔZPE

without additional water molecule NH3 + H 2O ⇌ NH3·H 2O

reaction

−5.9

−4.9

−2.4

CO2 + H 2O ⇌ CO2 ·H 2O

−2.6

−2. 6

−1.2

NH3 + CO2 ⇌ NH3·CO2

−2.9

−2.7

−1.7

NH3·H 2O + CO2 ⇌ NH3·H 2O·CO2

−4.6

−4.1

−2.7


−7.9

−6.4

−3.9

NH3·CO2 + H 2O ⇌ NH3·H 2O·CO2 NH3 + H 2O + CO2 ⇌ NH3·H 2O·CO2

NH3·H 2O·CO2 ⇌ TS1 TS1 ⇌ NH4HCO3

−7.6

−6.3

−3.4

−10.5

−9.0

−5.1

30.4

31.8

25.3

−19.8

−18.1

−17.3

NH3·H 2O + NH3 ⇌ (NH3)2 ·H 2O

−4.1

−1.9

0.1

NH3·CO2 + NH3 ⇌ (NH3)2 ·CO2

−4.7

−3.6

−1.8

(NH3)2 ·H 2O + CO2 ⇌ (NH3)2 ·H 2O·CO2

−6.3

−6.5

−5.3

(NH3)2 ·CO2 + H 2O ⇌ (NH3)2 ·H 2O·CO2

−8.7

−7.0

−4.0

NH3·H 2O·CO2 + NH3 ⇌ (NH3)2 ·H 2O·CO2

−6.0

−4.4

−2.5

(NH3)2 ·H 2O·CO2 ⇌ TS2

24.6

27.0

19.0

−20.3

−19.1

−16.8

−10.2

−6.2

−3.8

28.1

29.0

20.7

TS1·H 2O ⇌ (NH4HCO3)·H 2O

−18.9

−17.7

−15.9

(NH3)2 ·H 2O·CO2 + H 2O ⇌ ((NH3)2 ·H 2O·CO2 )·H 2O

−11.1

−7.6

−4.3

24.9

25.9

16.8

−19.8

−19.1

−17.0

TS2 ⇌ (NH4)2 CO3 with one additional water molecule

NH3·H 2O·CO2 + H 2O ⇌ (NH3·H 2O·CO2 )·H 2O

(NH3·H 2O·CO2 )·H 2O ⇌ TS1·H 2O

((NH3)2 ·H 2O·CO2 )·H 2O ⇌ TS2·H 2O TS2·H 2O ⇌ ((NH4)2 CO3)·H 2O

respect to reactants for each of the steps in each pathway is given in Table 4, and summarized in Figure 5. The calculated vibrational frequencies associated with each of the optimized species and their vibrational mode assignments are provided in Table 5 as Supporting Information. The optimized structure of each complex species is determined to be a local minimum configuration on the potential energy surface along the reaction pathway, with all frequencies being positive. The lowest frequency of these complexes was calculated to be in the range of 15−61 cm−1, featuring either the rocking or the twisting vibrational motions of the complex. The optimized transition state structures, TS1 and TS2, are saddle points on the potential energy surface of reactions 1 and 2, respectively, as indicated by one imaginary frequency of 608i cm−1 and 706i cm−1, respectively, along the reaction coordinate. The transition state frequency for a hydrogen transfer process would be typically close to 1700i cm−1. The exceptionally low imaginary frequencies of TS1 and TS2 reflect the fact that, instead of describing only one hydrogen transfer process, these transition states involved the migration of several atoms to rearrange the complexes. Specifically, it is the hydrogen transfer coupled with simultaneous movements of other atoms, including oxygen and carbon, that results in the low imaginary frequency for these transition states. IRC calculations were also performed to ascertain the connection between these transition states and their corresponding reactants

and products in reactions 1 and 2. It can be seen from Figure 4 that the bond lengths, r(CO), r(N−H), and r(O−H), were calculated to be 1.169 Å, 1.014 Å, and 0.958 Å for CO2, NH3, and H2O molecules, respectively, at the MP2/6-311G(d,p) level of theory. These values are in excellent agreement with the corresponding values of 1.162 Å, 1.012 Å, and 0.957 Å that are experimentally determined.20 The calculated bond angles of 106.0° and 102.4° for the NH3 and H2O molecules also bear significant resemblance to the experimental values of 106.7° and 104.5°, respectively.20 The good agreement between the computational structures and experimental measurements for these molecules suggests that the MP2/6-311G(d,p) level of theory is a reasonable method for structural prediction of the reactants, products, and intermediate species involved in reactions 1 and 2. The interaction between a water molecule and ammonia molecule is predicted to be of a hydrogen-bond type. The complex is formed by a hydrogen atom of the water molecule pointing to the nitrogen atom of the ammonia molecule (Figure 4 IV). A bond length of 1.966 Å for these two atoms leads to an internuclear distance of 2.935 Å between the oxygen and the nitrogen atoms, signifying the hydrogen-bond nature of this interaction.21 The r(N−H) bond lengths are essentially unaltered after the NH3 complexes with the water molecule. One of the r(O−H) bond lengths, on the other hand, is predicted 8995



Article

Figure 5. Calculated relative energy (in kcal mol−1) for the reaction of CO2(g) + nH2O(g) + nNH3(g) → (NH4)HCO3/(NH4)2CO3(s) (n = 1 and 2) at the MP4SDTQ/6-311++G(d,p)//MP2/6-311G(d,p)+ΔZPE level of theory (nonparenthesized values) and using the CBS-QB3 method (parenthesized values). Panel a presents the relative energy without the presence of an additional water molecule (n = 1), and panel b presents the relative energy with the presence of an additional water molecule (n = 2).

to increase slightly from 0.958 Å to 0.969 Å, indicating the influence of the formation of the hydrogen bond on the water molecular structure. Our optimized NH3·H2O complex structure is in good agreement with previous MP2 and DFT computational results.22 The interaction between a water molecule and carbon dioxide also leads to formation of a complex distinct from the NH3·H2O complex. The CO2·H2O complex features a bond length of 2.776 Å between the carbon atom of the CO2 molecule and the oxygen atom of the H2O molecule (Figure 4 V). While the r(O−H) and r(C−O) bond lengths do not change, the bond angle of water slightly increases and that of CO2 decreases in the predicted complex configuration. These parameters are consistent with previous computational results.23−26

The calculation results in the present work also indicate that the interaction between an ammonia molecule and a carbon dioxide molecule can yield a NH3·CO2 complex, in which the carbon atom of CO2 weakly connects to the nitrogen atom of NH3, with a C−N bond length of 2.902 Å (Figure 4 VI). A NH3·H2O·CO2 complex (Figure 4 VII) is expected to form by addition of a CO2 molecule to the NH3·H2O complex, addition of a NH3 molecule to the CO2·H2O complex, or addition of a H2O molecule to the NH3·CO2 complex. The NH3·H2O·CO2 complex has similar structural elements of both NH3·H2O and CO2·H2O complexes in terms of the N···H−O hydrogen bond and CO2 positioning relative to H2O. However, the water molecule is tilted toward the NH3 molecule to form the hydrogen bond, as evidenced by the ∠(7H−5O−8C) angle. The r(2N···7H) is decreased by 0.046 Å compared to that of 8996



Article

NH3·H2O, and the r(O−C) is decreased by 0.028 Å compared to that of CO2·H2O, suggesting that the NH3·H2O·CO2 complex is more compact than the individual NH3·H2O and CO2·H2O complexes. This could be due to the interaction between the 4H atom on NH3 and the 9O atom on CO2, with a r(9O−4H) of 2.454 Å. The (NH4)HCO3 molecule is calculated to be the product of a hydrogen-migrating rearrangement reaction of the NH3·H2O·CO2 complex. The optimized (NH4)HCO3 molecular structure suggests that all atoms except two hydrogen atoms (1H and 3H) are essentially in the same plane, as indicated by the dihedral angles of the molecule (Figure 4 VIII). It also suggests that this molecule structurally consists of an NH3 molecule linked to a H2CO3 molecule by a hydrogen bond, with a bond length of r(2N···4H) = 1.726 Å. Relative to the NH3·H2O·CO2 complex, the rearrangement reaction leads to a decrease of the r(5O−8C) bond from 2.748 Å to 1.375 Å, and an increase of the r(9O−8C) bond from 1.171 Å to 1.318 Å, which is in resonance with the r(5O−8C) bond. Meanwhile, the ∠(9O−8C−10O) angle is decreased from 176.5° to 126.0°. A transition state, TS1 (Figure 4), was located for reaction 1 in the present work, and the IRC computational result confirmed that the TS1 is the saddle point on the potential energy surface connecting the reactant, NH3·H2O·CO2, to the product, (NH4)HCO3. The TS1 has a six-member ring configuration composed of two hydrogen atoms (4H and 7H), two oxygen atoms (5O and 9O), one carbon atom (8C) and one nitrogen atom (2N), with r(7H−5O), r(7H−2N), r(4H−2N), r(4H− 9O), and r(8C−5O) bond lengths between those of NH3·H2O·CO2 and (NH4)HCO3. The corresponding angles of ∠(9O−8C−10O), ∠(9O−8C−5O), ∠(4H−9O−8C), and ∠(7H−5O−8C) are also between that of NH3·H2O·CO2 and (NH4)HCO3. Our IRC calculation results indicate that as the 7H atom moves away from the 5O atom toward the 2N atom to form a new N−H bond, the 4H atom moves away from the 2N atom toward the 9O atom to form a new O−H bond, and the 5O atom moves toward the 8C atom to form a new C−O bond. It is the migration of the 7H and 4H atoms in the NH3·H2O·CO2 complex that causes the relocation of these hydrogen atoms, leading to the formation of the (NH4)HCO3 product. The NH3 molecule thus plays a vital role as a medium in this process by accepting the hydrogen atom from the water molecule and donating a hydrogen atom to the CO2 molecule, yielding the H2CO3 component of the (NH4)HCO3 molecule. Our calculation also predicts that (NH 3 ) 2 ·H 2 O or (NH3)2·CO2 complexes can form when two NH3 molecules interact with either a water or a carbon dioxide molecule (Figure 4 IX and X). The (NH3)2·H2O complex has one hydrogen bond of r(N2···H7) = 1.975 Å, and both its r(O−H) bond lengths are slightly greater than that of the water molecule. The bond angles, ∠(2N−7H−5O) and ∠(9N−6H−5O), are predicted to be 172.9° and 130.5°, respectively. The asymmetric configuration of (NH3)2·H2O reflects the repulsion between two NH3 molecules connected to the same water molecule. Although the optimized (NH3)2·CO2 complex structure indicates that adding an ammonia molecule to the NH3·CO2 complex creates two hydrogen bonds via interactions between 11H and 2N, and 4H and 6O, with bond lengths of r(11H···2N) = 2.199 Å and r(4H···6O) = 2.381 Å, respectively, the structure of the NH3·CO2 complex remains essentially unaltered. Since ammonium carbamate has already been detected by Li et al., 17 the present work has attempted to optimize a NH2COONH4 species as a stationary minimum on the potential

energy surface. However, this attempt was unsuccessful, suggesting that NH2COONH4 might not exist as a single gaseous compound. Although (NH 3 ) 2 ·CO 2 , and not NH2COONH4, was located as an optimized gaseous phase species, it is likely that the (NH3)2·CO2 complex is transformed into NH2COONH4 as multiple (NH3)2·CO2 units coagulate to form a cluster, similar to how NH3·HNO3 is transformed into NH4+NO3− as four NH3·HNO3 units coagulate.27 While these calculations were not pursued in the present work due to the limitations of our computational resources, further investigations into these chemical systems is necessary to fully understand the experimental observations. Addition of a CO2 molecule to the (NH3)2·H2O complex, a NH3 molecule to the NH3·H2O·CO2 complex, or a H2O molecule to the (NH3)2·CO2 complex produces another complex, (NH3)2·H2O·CO2, shown as XI in Figure 4. The (NH3)2·H2O·CO2 complex is predicted to hold two N−H type hydrogen bonds formed between 12N and 6H, and 2N and 7H, with bond lengths of r(12N···6H) = 1.998 Å and r(2N···7H) = 1.997 Å, respectively. These N−H type hydrogen bonds are slightly longer in length than those found in the NH3·H2O·CO2 complex. The r(5O−8C) bond length is calculated to be 2.686 Å, which is slightly shorter than that of the NH3·H2O·CO2 complex. Rearrangement of the (NH3)2·H2O·CO2 complex results in the formation of (NH4)2CO3, and the optimized structure of (NH4)2CO3 is displayed in Figure 4 XII. Similar to (NH4)HCO3, the (NH4)2CO3 molecule can be regarded as a complex made of two NH3 molecules and one H2CO3 molecule, and all atoms except 1H, 3H, 11H, and 14 H are expected to be in the same plane. However, the N−H hydrogen bond lengths of this complex are greater than those of (NH4)HCO3 by about 0.24 Å. The r(5O−8C) bond length is shorter, but the other C−O bond lengths of (NH4)2CO3 are slightly longer than those of (NH4)HCO3. Note that the (NH4)2CO3 also appears to be structurally less symmetric than the (NH3)2·H2O·CO2 complex. This may be due to the decreased ability of atom 5O to attract electrons because of involvement in the 5O−8C−10O resonance structure, causing a longer 5O−6H bond length than the 9O−4H bond length. The transition state structure for the rearrangement of the (NH3)2·H2O·CO2 complex to (NH4)2CO3 is given as TS2 in Figure 4. Our IRC calculation result indicates that TS2 is the configuration of a saddle point on the potential energy surface for reaction 2. This structure also involves a six-member ring character for the simultaneous migrations of a hydrogen atom (6H) from the water molecule to the ammonia molecule and a hydrogen atom (13H) from ammonia molecule to the carbon dioxide molecule. The hydrogen atom 6H is expected to position 1.298 Å and 1.188 Å away from the water molecule and ammonia molecule, and the hydrogen atom 13H is calculated to separate from the carbon dioxide molecule by 1.708 Å, respectively. These TS2 parameters are comparable to the corresponding parameters of 1.343 Å, 1.157 Å, and 1.734 Å in TS1. Again, one of the ammonia molecules in the TS2 serves as the medium for the simultaneous acceptance and donation of hydrogen atoms to facilitate the conversion of (NH3)2·H2O·CO2 complex into (NH4)2CO3. Meanwhile the remaining ammonia molecule complexes with both the H2O and CO2 molecules. Finally, the r(5O−8C) bond is significantly shorter in TS2 than in the (NH3)2·H2O·CO2 complex, indicating a more “product-like” feature of this transition state. The calculated total energy associated with each of the species in Figure 4 is listed in Table 3, and the relative energy values of 8997



Article

under flow conditions while the experiments of the present work were performed under static conditions, and the reactants experienced a different reaction times in the reactors. The reaction time in the flow reactor of ref 17 was in the range of 47− 188 s, while the reaction time in the present work was on the order of 10 s after a pulse of NH3 was injected into the reactor (Figure 2). A long residence time may have allowed secondary reactions to occur, converting (NH4)2CO3 into NH4HCO3 via

reactions 1 and 2 are given in Table 4. Total energy calculated at the MP4/6-311++G(d,p) level of theory is lower than that calculated at the MP2/6-311G(d,p) level of theory due to both the use of a larger basis set and the inclusion of higher order electron interactions. Therefore, with zero point energy (ZPE) correction, the energy values calculated at the MP4/6-311+ +G(d,p)+ΔZPE level of theory are used as our best estimate for evaluation the energy changes for reactions 1 and 2. Our computational results suggest that the complex formations are all exothermic processes, with releasing an amount of heat of 4.9, 2.7, and 2.6 kcal mol−1 for the formation of NH3·H2O, NH3·CO2, and CO2·H2O from interactions between two species of H2O, NH3, and CO2, respectively. The complexations of NH3·H2O with CO2, NH3·CO2 with H2O, and CO2·H2O with NH3 to form the NH3·H2O·CO2 complex are predicted to be exothermic by 4.1, 6.3, and 6.4 kcal mol−1, respectively. However, the rearrangement of NH3·H2O·CO2 to NH4HCO3 is expected to be endothermic by 13.7 kcal mol−1, with an activation energy barrier of 31.8 kcal mol−1. Our calculation results also suggest that further complexation of NH3·H2O with NH3 to form the (NH3)2·H2O complex will release 1.9 kcal mol−1 of energy. On the other hand, addition of an NH3 to NH3·CO2 to form the (NH3)2·CO2 complex is expected to be exothermic by 4.1 kcal mol−1. The formation of the (NH3)2·H2O·CO2 complex may release 6.6, 7.0, and 4.4 kcal mol−1 of energy from further complexation of (NH3)2·H2O with CO2, (NH3)2·CO2 with H2O, and NH3·H2O·CO2 with NH3, respectively. Finally, the conversion of the (NH3)2·H2O·CO2 complex into (NH4)2CO3 via rearrangement is predicted to be endothermic by 7.8 kcal mol−1, and is required to overcome an activation energy barrier of 27.0 kcal mol−1. Figure 5a summarizes the calculated relative energy of each species involved in reactions 1 and 2 using the energy of the reactants, i.e., 2NH3 + H2O + CO2, as a reference. Relative to the reactants, the complex formation processes are all exothermic. Additionally, the pathway leading to the formation of NH3·H2O complex appears to be more exothermic relative to those leading to the formation of both the H2O·CO2 and NH3·CO2 complexes. Likewise, the pathways leading to the formation of NH3·H2O·CO2 complex appear to be more exothermic than the pathways leading to the formation of both (NH3)2·H2O and (NH3)2·CO2 complexes. As shown in Figure 5a, the formation of NH4HCO3 from NH3 + H2O + CO2 is endothermic by 4.7 kcal mol−1, and the production of (NH4)2CO3 from 2NH3 + H2O + CO2 is exothermic by 5.5 kcal mol−1. Further, at least 22.8 and 13.6 kcal mol−1 of energy are required to drive reactions 1 and 2 to produce NH4HCO3 and (NH4)2CO3, respectively. This suggests that while the reaction of CO2 with NH3 and H2O thermodynamically favors the generation of (NH4)2CO3 over NH4HCO3 in the gaseous phase, these chemical processes may not be kinetically favored. As a result, more (NH4)2CO3 should be generated than NH4HCO3, and both (NH4)2CO3 and NH4HCO3 are expected to be produced in small amounts over a long period of time when NH3, H2O, and CO2 are mixed in the reactor. These calculation results contrast with our experimental observations displayed in Figure 2. One possible explanation to reconcile our calculation results with the experimental observation is presented in the last paragraph of section c. It is also noticed that (NH4)2CO3(s) was detected as one of the final products in the present work, but not in ref 17. The cause of these different experimental observations is unclear at this time. One factor potentially contributing to this difference may be the fact that the experiments in ref 17 were carried out

(NH4)2 CO3(s/g) + CO2 (g) + H 2O(g) → 2NH4HCO3(s/g)

Such a process could explain why only NH2COONH4(s) and NH4HCO3(s) were observed in the experiments carried out in ref 17. Regarding the observation of (NH4)2CO3(s) being formed in the present study, it has been proposed that (NH4)2CO3(s) can be produced via17 NH3(g) + H 2O(l) ⇌ NH4OH(l) 2NH4OH(l) + CO2 (g) ⇌ (NH4)2 CO3(s) + H 2O(l)

This reaction mechanism can account for (NH4)2CO3(s) formation in ammonia scrubbing for removal of CO2 in aqueous phase, but may not be able to explain the formation of (NH4)2CO3(s) in the present work for two reasons. First, on the basis of our calculation results, an ammonium hydroxide molecule prefers to exist as a NH3·H2O complex over its NH4OH form in the gaseous phase. Second, the next step of this mechanism would be a ter-molecular reaction if it occurred in the gaseous phase, which would not be an efficient process. It is therefore more feasible for the (NH4)2CO3(s) to be produced in gaseous phase through a pathway involving the formation and rearrangement of the (NH3)2·H2O·CO2 complex. c. Theoretical Results: With the Presence of One Additional Water Molecule. Due to the presence of water molecules in this chemical system, it is likely that the additional water molecules may play a role in driving reactions 1 and 2 by interacting with the complexes formed in these reactions. The rearrangement reactions of NH3·H2O·CO2 to NH4HCO3 and of (NH3)2·H2O·CO2 to (NH4)2CO3 have been further studied theoretically in the present work with the presence of one additional water molecule to understand whether and how this additional water would affect the relative energy of reactions 1 and 2. The optimized geometry of NH3·H2O·CO2, TS1, NH4HCO3, (NH3)2·H2O·CO2, TS2, and (NH4)2CO3 in the presence of an additional water molecule is given in Figure 4 (rows 6 and 7). Our computational results suggest that the additional water molecule complexes with the NH3·H2O·CO2, TS1, and NH4HCO3 by forming additional hydrogen bonds with the H2O molecule, resulting in formation of (NH3·H2O·CO2)·H2O (XIII), TS1·H2O, and (NH4HCO3)·H2O (XIV) complexes. In the case of (NH 3 ·H 2 O·CO 2 )·H 2 O, compared to the NH3·H2O·CO2 complex, the r(2N···7H) hydrogen bond is decreased by 0.076 Å, but the r(5O−8C) bond is increased by 0.209 Å. The distances of r(12O−6H) and r(10O−11H) are predicted to be 1.769 Å and 1.815 Å, respectively, in TS1·H2O, and 1.766 Å and 1.921 Å, respectively, in (NH4HCO3)·H2O, indicating an increasing interaction between these oxygen and hydrogen atoms as (NH3·H2O·CO2)·H2O transforms into (NH4HCO3)·H2O. Compared to TS1, the hydrogen atom 7H is closer to the oxygen atom 5O in TS1·H2O, and the TS1·H2O 8998



Article

has a longer r(9O−4H) bond (1.816 Å), and a shorter r(5O− 8C) bond (1.537 Å). Similarly, addition of a water molecule to (NH3)2·H2O·CO2, TS2, and (NH 4 ) 2 CO 3 leads to the formation of ((NH3)2·H2O·CO2)·H2O (XV), TS2·H2O, and ((NH4)2CO3)·H2O (XVI), respectively. Our computational results indicate that the additional water molecule complexes to (NH3)2·H2O·CO2 by forming a new hydrogen bond, r(5O···17H), with a bond length of 1.845 Å. Compared to (NH3)2·H2O·CO2, this complexation results in a slight decrease in the r(2N···7H) (1.928 Å) and r(12N···6H) (1.918 Å) hydrogen bond lengths, but a small increase in the r(10O−13H) (2.567 Å) and r(5O−8C) (2.697 Å) distances. During the rearrangement reaction, the r(5O···17H) hydrogen bond is lost due to the rotational motion of the hydrogen atom 17H around the 16O−15H bond as the NH3 group consisting of 2N, 1H, 3H, and 4H moves toward the oxygen atom 9O. At the end of the rearrangement, this water molecule complexes to the (NH4)2CO3 molecule via the r(9O···15H) hydrogen bond, with a bond length of 1.832 Å. As in TS2, the six-membered ring of the transition state structure remains in TS2·H2O, and the majority of its bond lengths for the migrating hydrogen atoms are only slightly altered relative to TS2. Finally, compared to (NH4)2CO3, the additional water molecule causes both the r(2N−7H) hydrogen bond and the r(5O−8C) bond to decrease by 0.063 Å and 0.013 Å, respectively. The total energy of each species associated with the additional water molecule is listed in Table 3, and the energy values of the reactions that include the additional water molecule are also given in Table 4 and summarized in Figure 5b. The enthalpy of the overall reaction involving reactions 1 and 2 calculated using the CBS-QB3 method is comparable to that calculated at the MP4SDTQ/6-311++G(d,p)//MP2/6-311G(d,p)+ΔZPE level of theory. The difference between these calculations is less than 2 kcal mol−1. These calculations suggest that the MP4SDTQ single-point calculation can be used as an approximation for the best estimate of the thermodynamic property for such reactions. It is found that the presence of the additional water molecule significantly changes the relative energy of reactions 1 and 2. The complexation of the additional water molecule to the primary complex contributes an additional 6.2 kcal mol−1 and 7.6 kcal mol−1 to the exothermicity for reactions of 1 and 2, respectively. Our computational results also show that at the MP4SDTQ/6311++G(d,p)//MP2/6-311G(d,p)+ΔZPE level of theory, the presence of one additional water molecule lowers the activation energy barrier by 2.8 kcal mol−1 and 1.0 kcal mol−1 for the rearrangement of NH3·H2O·CO2 into NH4HCO3 and of (NH3)2·H2O·CO2 into (NH4)2CO3, respectively. As illustrated in Figure 5b, the presence of one additional water molecule significantly changes the potential energy surface for reactions 1 and 2. Compared to Figure 5a, the presence of an additional water molecule causes the formation of NH4HCO3 and (NH4)2CO3 to be exothermic by 3.9 kcal mol−1 and 14.0 kcal mol−1, respectively, and requires less overall energy (13.8 kcal mol−1 and 5.1 kcal mol−1, respectively) to drive these reactions. The increase in exothermicity and decrease in activation energy ultimately facilitate the reaction of the CO2 with NH3 and H2O to produce the NH4HCO3 and (NH4)2CO3. Nonetheless, the high activation energy calculated at the MP4SDTQ level of theory suggests that reactions 1 and 2 are still not kinetically favored. The rapid formation of the NH4HCO3 and (NH4)2CO3 products could thus be attributed to the presence of high concentrations of reactants in the reactor since

the rates of reactions 1 and 2 depend on the reactant concentrations. In particular, the high concentrations of [H2O] > 1 × 1017 molecules cm−3 and [CO2], [NH3] > 1 × 1018 molecules cm−3 might have contributed to the formation of both the NH4HCO3 and the (NH4)2CO3 products observed in the present work. It is plausible that when more water molecules are present in the reactor, the potential energy surface could be further changed to favor these reactions due to the additional complexations the water molecules instigate. It is also likely that when excess water molecules are available, NH3 hydration may play a role in reducing the activation energy of reactions 1 and 2. Finally, the clustering process converting NH4HCO3(g) and (NH4)2CO3(g) into NH4HCO3(s) and (NH4)2CO3(s) is expected to be exothermic, similar to the process involving the conversion of NH4NO3(g) to NH4NO3(s).27 The heat released from such coagulation processes may subsequently drive the gaseous phase portions of reactions 1 and 2. However, further computational studies are needed to verify this supposition.

■

CONCLUSION This study shows that carbon dioxide reacts with ammonia and water to produce ammonium bicarbonate and ammonium carbonate. Ab initio calculations suggest that the gaseous phase reaction may proceed through the complex formation of either NH3·H2O, NH3·CO2, or H2O·CO2, followed by formation of the NH3·H2O·CO2 complex by interaction of either NH3·H2O with CO2, NH3·CO2 with H2O, or H2O·CO2 with NH3. This NH3·H2O·CO2 formation pathway is more exothermic than the pathway leading to the formation of either (NH3)2·CO2 or (NH3)2·H2O by addition of an ammonia molecule to either NH3·CO2 or NH3·H2O. Our calculations suggest that the NH3·H2O·CO2 complex is weakly held together by the N−H type hydrogen bond plus the interaction between hydrogen and oxygen atoms. The rearrangement of NH3·H2O·CO2 leads to the formation of NH4HCO3. (NH4)2CO3 can be produced from the rearrangement of the (NH3)2·H2O·CO2 complex that is produced either by adding a NH 3 molecule to the NH3·H2O·CO2 complex, by adding CO2 to the (NH3)2·H2O complex, or by adding H2O to the (NH3)2·CO2 complex. Therefore, complexation may play a key role in these chemical processes as precursors for the final products. N−H and/or O− H type hydrogen bonds are found in all complexes except H2O·CO2 and NH3·CO2, and these hydrogen bonds serve to hold the molecules together. The rearrangement reactions of the NH3·H2O·CO2 complex into NH4HCO3 and the (NH3)2·H2O·CO2 complex into (NH4)2CO3 are found to involve simultaneous migration of two hydrogen atoms, one from the water molecule to the ammonia molecule and another from the ammonia molecule to the carbon dioxide molecule. Hence the NH3 molecule acts as a medium that accepts and donates hydrogen atoms in reactions 1 and 2. Our theoretical calculation results also indicate that the rearrangement reactions can be accelerated in the presence of one additional water molecule, which increases the exothermicity and decreases the activation energy of the overall reaction by complexing to the previously formed complexes. However, the predicted overall activation energy of 5.1−13.8 kcal mol−1 suggests that reactions 1 and 2 are expected to be slow processes, but rapid formation of NH4HCO3 and (NH4)2CO3 products can be achieved when the reactants’ concentrations are high. The clustering process of the final products may release heat to further facilitate the gaseous phase reactions 1 and 2 in the reactor. This proposed process may account for the formation of 8999



Article

the NH4HCO3(s) and (NH4)2CO3(s) final products in the form of aerosol particles, as observed in our experiments. In spite of the significant obstacles associated with removing CO2 from the atmosphere, the reaction of CO2 with NH3 and H2O may be considered as an alternative way to reduce CO2 emissions from point sources, such as power plants and cement plants, which may potentially help reduction of global warming. The NH4HCO3 and (NH4)2CO3 produced from this reaction may be used as acidity regulators and raising agents, and NH4HCO3 has been used as fertilizer in China. However, more studies are needed to understand the kinetics of this reaction and the optimal ratio of the reactants to establish environmental engineering design for most efficient removal of the CO2.

■

(15) Song, C. Catal. Today 2006, 115, 2−32. (16) Mandal, B. P.; Bandyopadhyay, S. S. Chem. Eng. J. 2006, 61, 5440−5447. (17) Li, X.; Hagaman, E.; Tsouris, C.; Lee, J. W. Energy Fuels 2003, 17, 69−74. (18) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N.; Burant, J. C.; et al. Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2010. (19) Meng, L.; Burris, S.; Bui, H.; Pan, W. P. Anal. Chem. 2005, 77, 5947−5952. (20) Herzberg, G. Molecular Spectra and Molecular Structure: Electronic Spectra and Electronic Structure of Polyatomic Molecules, 2nd ed.; Krieger Publishing Company: Malabar, FL, 1966; Vol. III. (21) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry, 4th ed; John Wiley & Sons, Inc.: New York, 1980. (22) Sadlej, J.; Moszynski, R.; Dobrowolski, J. Cz.; Mazurek, A. P. J. Phys. Chem. A 1999, 103, 8528−8536. (23) Danten, Y.; Tassaing, T.; Besnard, M. J. Phys. Chem. A 2005, 109, 3250−3256. (24) Garden, A. L.; Lane, J. R.; Kjaergaard, H. G. J. Chem. Phys. 2006, 125, 144317−144324. (25) Makarewicz, J. J. Chem. Phys. 2010, 132, 234305−234315. (26) Xu, Z.; Sander, S. P. J. Phys. Chem. A 2011, 115, 9854−9860. (27) Zhang, B.; Tao, F.-M. Chem. Phys. Lett. 2010, 489, 143−147.

ASSOCIATED CONTENT

S Supporting Information *

The calculated frequencies and zero point energy of the species involved in the title reaction (Table 5) are available as Supporting Information. This information is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: (657)278-3585. Fax: (657)278-8010. E-mail: zli@ fullerton.edu. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is supported in part by the National Science Foundation (NSF ATM-0533574) and by the CSUF University Mission and Goal Initiative (UMGI). The authors thank Dr. F.M. Tao for his helpful discussion. Z.L. would also like to thank Mr. Brad van Mourik for his assistance in the computer operations.

■

REFERENCES

(1) Tans, P. Oceanography 2009, 22 (4), 26−35. (2) Friedlingstein, P.; Solomon, S. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 14256−14261. (3) Solomon, S.; Plattner, G.-K.; Knutti, R.; Friedlingstein, P. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 1704−1709. (4) Matthews, H. D.; Gillett, N. P.; Stott, P. A.; Zickfeld, K. Nature 2009, 459, 829−833. (5) Strand, S. E.; Benford, G. Environ. Sci. Technol. 2009, 43, 1000− 1007. (6) Bonenfant, D.; Kharoune, L.; Sauvé, S.; Hausler, R.; Niquette, P.; Mimeault, M.; Kharoune, M. Ind. Eng. Chem. Res. 2008, 47, 7617−7622. (7) Siriwardane, R.; Robinson, C. J. Environ. Eng. - ASCE 2009, June, 378−385. (8) Chang, H.; Wu, Z. X. Ind. Eng. Chem. Res. 2009, 48, 4466−4473. (9) Bai, H.; Yeh, A. C. Ind. Eng. Chem. Res. 1997, 36, 2490−2493. Brooks, R. Manufacture of Ammonium Bicarbonate. British Patent 742,386, 1953. Mani, F.; Peruzzini, M.; Stoppioni, P. Green Chem. 2006, 8, 995−1000. (10) Mani, F.; Peruzzini, M.; Stoppioni, P. Energy Fuels 2008, 22, 1714−1719. (11) Mani, F.; Peruzzini, M.; Barzagli, F. ChemSusChem 2008, 1, 228− 235. (12) Nikulshina, V.; Gebald, C.; Steinfeld, A. Chem. Eng. J. 2009, 146, 244−248. (13) House, K. Z.; House, C.; Schrag, D. P.; Aziz, M. J. Environ. Sci. Technol. 2007, 41, 8464−8470. (14) Sun, P.; Grace, J. R.; Lim, C. J.; Anthony, E. J. Energy Fuels 2007, 21, 163−170. 9000


Experimental and Theoretical Investigation of Homogeneous Gaseous

Recommend Documents