13C and 15N NMR characterization of amine reactivity and solvent effects in CO2 capture

Article abstract of DOI:10.1021/jp503421x

Factors influencing the reactivity of selected amine absorbents for carbon dioxide (CO₂) capture, in terms of the tendency to form amine carbamate, have been studied. Four linear primary alkanolamines at varying chain lengths (MEA, 3A1P, 4A1B, and 5A1P), two primary amines with different substituents in the β-position to the nitrogen (1A2P and ISOB), a secondary alkanolamine (DEA), and a sterically hindered primary amine (AMP) were investigated. The relationship between the ¹⁵N NMR data of aqueous amines and their ability to form carbamate, as determined at equilibrium by quantitative ¹³C NMR experiments, was analyzed, taking into account structural-chemical properties. For all the amines, the ¹⁵N chemical shifts fairly reflected the observed reactivity for carbamate formation. In addition to being a useful tool for the investigation of amine reactivity, ¹⁵N NMR data clearly provided evidence of the importance of solvent effects for the understanding of chemical dynamics in CO₂ capture by aqueous amine absorbents.

Full text of DOI:10.1021/jp503421x

Article
pubs.acs.org/JPCB
15
13C and N NMR Characterization of Amine Reactivity and Solvent
Effects in CO2 Capture
Cristina Perinu,† Bjørnar Arstad,‡ Aud. M. Bouzga,‡ and Klaus-J. Jens*,†
†Faculty of Technology, Telemark University College, Kjølnes Ring 56, 3901 Porsgrunn, Norway
‡SINTEF Materials and Chemistry, Forskningsveien 1, 0314 Oslo, Norway
S
* Supporting Information
ABSTRACT: Factors influencing the reactivity of selected amine absorbents for carbon dioxide
(CO2) capture, in terms of the tendency to form amine carbamate, have been studied. Four linear
primary alkanolamines at varying chain lengths (MEA, 3A1P, 4A1B, and 5A1P), two primary
amines with different substituents in the β-position to the nitrogen (1A2P and ISOB), a
secondary alkanolamine (DEA), and a sterically hindered primary amine (AMP) were
investigated. The relationship between the 15N NMR data of aqueous amines and their ability
to form carbamate, as determined at equilibrium by quantitative 13C NMR experiments, was
analyzed, taking into account structural−chemical properties. For all the amines, the 15N
chemical shifts fairly reflected the observed reactivity for carbamate formation. In addition to
being a useful tool for the investigation of amine reactivity, 15N NMR data clearly provided
evidence of the importance of solvent effects for the understanding of chemical dynamics in CO2
capture by aqueous amine absorbents.
related to the ability of an amine to form carbamate, which
depends on chemical-structural properties of the amine and
reaction and process conditions; and this is the reason why the
carbamate formation step is considered to be a distinguishing
factor between the amines.6 Identification of the factors
influencing the tendency of an amine to form carbamate is
therefore important for improving the CO2 absorption
processes.
1. INTRODUCTION
Acid gas (e.g., CO2, H2S) scrubbing by chemical absorption
into aqueous alkanolamine solutions is widely practiced in the
gas industry and may become the first deployed technology for
postcombustion carbon capture (PCC) for global warming
abatement.1 The most widely used solvent for these purposes is
aqueous monoethanolamine (MEA), the benchmark absorbent,
which is known for its high reactivity and favorable reaction
kinetics toward CO2, although degradation and unfavorable
thermodynamics reduce its potential for being the ideal PCC
absorbent for energy industry applications.2
In gas processing, physical dissolution of CO2 into the water
phase takes place before the reaction of CO2 with an amine.
Primary and secondary amines react directly with CO2 to form
amine carbamate (in thermodynamic equilibrium with carbamic
acid for aqueous-phase reactions) (reaction 1).3 In contrast,
tertiary amines and some so-called sterically hindered amines
act as bases accepting a proton from the carbonic acid (a
product of CO2 with water) and/or possibly as catalysts in the
CO2 hydration (reaction 2).4
Several structure−activity relationship studies have discussed
the influence of electronic and steric effects, together with
chemical properties of amines in the reaction with CO2 in order
to obtain information on governing factors for solvent
performances. Recently, Yamada et al. investigated the CO2
absorption capacity of secondary alkanolamines with varied
alkyl and alcohol chain lengths by combining computational
data and 13C NMR experiments.7 As the distance between the
hydroxyl (−OH) and amino (−NH) functional groups
increased within the amine structure, the amount of carbamate
formed at equilibrium was decreased, and CO2 absorption
capacity was increased. In contrast, varied alkyl chain length did
not have a significant effect. The sensitivity to the alcohol chain
length was attributed to intramolecular hydrogen bonds
between −OH and −NH in neutral alkanolamines, −OH and
2R1R2NH + CO2 ⇆ R1R2NCOO− + R1R2NH2
+
(1)
R1R2R3N + CO2 + H2O ⇆ HCO3− + R1R2R3NH+
−NH2 in protonated alkanolamines, −OH and −NCOO− in
+
(2)
carbamate anions. However, the role played by intermolecular
hydrogen bonds was neither clear nor excluded.7 Puxty et al.
published a systematic screening study of the CO2 absorption
capacity of 76 structurally diverse amines, and seven of them
Reaction 2 is more efficient than reaction 1 in terms of CO2
absorption capacity, but reactions of primary and secondary
amines with CO2 (reaction 1) show the fastest reaction kinetics.
However, during the CO2 desorption/amine regeneration step,
the energy demand for the reverse of reaction 1 is higher than
reaction 2 due to the stability of the carbamates.4,5 Therefore,
the CO2 absorption capacity of an amine−CO2−H2O system is
Received: April 7, 2014
Revised: July 10, 2014
Published: August 5, 2014
© 2014 American Chemical Society
10167
dx.doi.org/10.1021/jp503421x | J. Phys. Chem. B 2014, 118, 10167−10174

The Journal of Physical Chemistry B
Article
were identified for outstanding activity.8 These seven amines
had some common structural features, such as the hydroxyl
group located two or three carbons distant from the amino
functionality, but the role played by this structural characteristic
was unclear.8 Furthermore, with respect to the amine basicity,
some Brønsted correlations relating rate and equilibrium
constants for the carbamate formation to protonation contants
of amines have been reported in the literature.6,9 Hamborg et
al. described the base strength of the alkanolamines to be
dependent on the dielectric constants and temperature of the
solvent.10 However, no clear trend between the CO2 absorption
capacity of the amines and the corresponding basicity was
identified.8,11
In view of the fact that water is the predominant component
in amine solvents for CO2 scrubbing, only a few reports have
considered the properties of water (e.g., high polarity and
extensive hydrogen bonding ability) in these reactions. Han et
al. reported that water could be considered as a spectator in the
reaction between amine and CO2;12 in contrast Arstad et al.
showed in a computational study that water molecules can act
as catalysts for the CN bonding in the formation of carbamic
acid (the corresponding acid to the carbamates)13 and da Silva
has taken into account solvation in terms of stabilizing effect
depending on structural accessibility.14 However, the influence
of the water solvent has never been considered in terms of the
availability of the nitrogen’s lone pair of electrons to interact
with water, although reactivity and CO2 absorption capacity of
the amines could be influenced greatly.
In the current study, we have measured the amount of
carbamate formed at equilibrium in reactions of different
−
amines with bicarbonate (HCO3 ), by means of quantitative
13C NMR experiments, and compared these values to 15N
NMR data and structural-chemical properties of the selected
amines.
We have examined linear primary alkanolamines with carbon
chains of varying length from two (2-amino-1-ethanol, termed
MEA or ethanolamine) to five methylenes (5-amino-1-
pentanol, 5A1P) between the hydroxyl and amino nitrogen
functional groups. Two other primary amines, 1-amino-2-
propanol (1A2P) and isobutylamine (ISOB), featuring the
same carbon chain length but a different substituent at the
position β to the nitrogen, were also analyzed to understand the
effect of the hydroxyl function on the amine structure.
Furthermore, 2,2′-iminodiethanol (also termed diethanolamine,
DEA) and 2-methyl-2-amino-1-propanol (AMP), a secondary
and a sterically hindered amine, respectively, were included in
the investigation to scrutinize overall structure−activity
relationships (Figure 1).
In order to provide further insight into amine solvents as
absorbents for CO2 capture, we have applied 13C and 15N NMR
to characterize the amine reactivity, in terms of tendency to
form amine carbamate. The background is that, during
carbamate formation, the amino nitrogen is acting as a
nucleophile (Lewis base) donating an electron pair to an
Figure 1. Amines investigated in this study.
This approach allowed us to identify overall factors
influencing the tendency of the selected amines to form
carbamate. In particular, 15N NMR spectroscopy was a useful
tool to investigate the amine reactivity toward formation of
amine carbamate, as determined by 13C NMR spectroscopy,
and to examine the role played by the solvent (e.g., water).
−
electrophile (Lewis acid), such as CO2 and/or HCO3 , and the
ability of a nucleophile to attack an electrophile depends not
only on chemical structural properties of the molecules but also
on medium effects. Increased electron density on the nitrogen
raises the energy of the electron pair and makes it more
reactive, which is the reason why the reactivity is thus strongly
influenced by the availability of the electron lone pair of the N
nucleus, making parameters describing the local electronic
properties on the N atoms important for understanding these
reactions. A technique that has been considered a useful tool to
assess the electron density on the amino nitrogen atom and to
study solvent interactions is 15N NMR spectroscopy, since it
can provide information about the lone pair availability of
nitrogen and the factors influencing the electron density on this
nucleus, directly through the measured chemical shift values.15
Indeed, 15N chemical shift values not only depend on the
electronic chemical environment defined by the molecular
structure but, as compared to 1H and 13C NMR, are also much
more sensitive to medium effects (e.g., concentration, temper-
ature, and solvent) and, in general, to inter- and intramolecular
interactions of the amino nitrogen with other functional
groups.15
2. EXPERIMENTAL SECTION
2.1. Sample Preparation. The following chemicals were
used in the present study: 2-Amino-1-ethanol (EMSURE) and
sodium hydrogen carbonate (sodium bicarbonate) from Merck,
3-amino-1-propanol (99%), 4-amino-1-butanol (98%), 5-
amino-1-pentanol (95%), isobutylamine (99%), (R/S)-1-
amino-2-propanol (98%), 2,2′-iminodiethanol (≥98%), 2-
amino-2-methyl-1-propanol (≥99%), ethanolamine hydrochlor-
ide (≥99%), and hydrochloric acid (37%) from Sigma-Aldrich.
They were utilized without any further purification.
Amines were weighed and solutions (2M) were prepared
with distillated and degassed water. The concentrations were
calculated by measuring the density with a pycnometer (5.554
cm3). The same procedure was used for preparation of water
(H2O)/dimethoxyethane (DME) (1:1) (2 M) amine solutions
(MEA and AMP) and for the aqueous protonated amines
(amineH+) and the 1:1 ratio amine/amineH+ (2 M) solutions
(MEA and 3A1P). In the first case, a weighted amount of amine
was dissolved in H2O−DME, previously mixed at 1:1 ratio; in
the second case, commercially available protonated MEA was
used, whereas protonated 3-amino-1-propanol was obtained by
adding equimolar amounts of hydrochloric acid (HCl) into the
amine solution. A 600 μL sample of the above solutions or of
Little focus has been given on 15N NMR within the field of
PCC, with the exception of Yoon et al., who reported a 15N
NMR study discussing the electronic effects of substituents in
sterically hindered amines on CO2 absorption capacity.
However, factors other than amine molecular structure (like,
e.g., hydrogen bonds and solvent effects) were not consid-
ered.16
10168
dx.doi.org/10.1021/jp503421x | J. Phys. Chem. B 2014, 118, 10167−10174

The Journal of Physical Chemistry B
Article
Figure 2. 15N NMR spectrum of aqueous MEA solution (2 M) at 298.15 K. The nitrogen is shown in bold in the formula.
Table 1. 15N Chemical Shift Values of the Amines Solutions Investigated in This Study
amines
15N chemical shift (ppm)
amines (2M) in H2O
pure amines
amines (2 M) in H2O/DME 1:1
amine/amineH+ 1:1 (2M) in H2O
amineH+ (2M) in H2O
1A2P
MEA
3A1P
4A1B
17.74
18.38
24.26
25.39
25.34
22.15
29.58
48.75
17.29
16.84
22.89
24.06
18.05
24.34
29.42
29.01
33.66
a
5A1P
ISOB
18.70
a
DEA
a
AMP
48.44
a
Solid at STP conditions.
Figure 3. 13C NMR spectrum and assignments for MEA/NaHCO3 1:1 ratio reaction mixture at the equilibrium. The observed carbons are given in
bold in the formulas; CH3CN is the reference.
the neat amines (directly withdrawn from the bottle) was
inserted in the NMR tube for 15N NMR measurements.
The synthesis of amine carbamate was carried out by reacting
the aqueous amine solutions (2 M) with sodium bicarbonate at
1:1 molar ratio (reaction 3). The mixtures were stirred for more
than 24 h at 298.15 K to achieve equilibrium and after 48 h
quantitative 13C NMR experiments were performed.
chemical shift values, the referencing via direct measurements
of the absolute frequency of the field/frequency lock signal was
used.17 A capillary containing deuterated benzene was inserted
in the NMR tube for locking and referencing and, in a separate
NMR tube, pure formamide (δ = 113.3 ppm) was used to
validate the ppm values.17 This method was applied to replace
medium effects on the shielding of the reference standard in
such solutions and reduce the acquisition time. Indeed, the
relatively low amount of standard reference that would be
added into the NMR tubes, combined with the low isotopic
abundance of 15N (0.37%), would result in long acquisition
time. Each aqueous amine solution (2M) was prepared twice
for 15N NMR measurements, and the uncertainty in the
chemical shift values was estimated to be in the range of
0.01−0.03 ppm.
RNH2 + HCO3− ⇆ RNHCOO− + H2O
(3)
2.2. NMR Experiments. 13C and 15N NMR experiments
were performed at 9.4 T on a Bruker Avance III 400 MHz
spectrometer using a BBFO Plus double resonance probe head
at 298.15 K; the spectra were processed using MestreNova
software v 7.1.1.
2.2.1. Qualitative 15N NMR Experiments. Qualitative 15N
NMR experiments were carried out on the amine solutions at
the same concentration (2 M) and at constant temperature
(298.15 K) in order to eliminate the corresponding influences
on the chemical shifts. At the same temperature, 15N NMR
experiments were also performed on the neat amines. For all
amines, the experiments were run with the inverse gated
decoupling method, pulse angle of 90° (14 μs pulse width) and
a pre-scan delay of 250 μs (optimized to reduce probe ringing).
The choice of the recycle delay was based on the need of
observing a signal at short experimental time. Therefore, for all
amine solutions, a recycle delay of 10 s and scans up to 4352
were set, except for 2 M 5A1P solution which required a recycle
delay of 50 s and 512 scans. In order to record the 15N NMR
A typical 15N NMR spectrum is reported in Figure 2, whereas
the 15N chemical shift values of all the amine solutions are
reported in Table 1. 15N NMR spectra of all the aqueous
amines solutions at 2 M can be found in the Supporting
Information (SI).
2.2.2. Quantitative 13C NMR Experiments. Quantitative 13
C
NMR experiments were performed on the equilibrated reaction
mixtures (aqueous solutions after carbamate formation).
Acetonitrile (CH3CN) and deuterated water (D2O), inserted
in a sealed capillary, were used as standard reference and lock
solvents, respectively.18 After the measurements of the
longitudinal relaxation time constant (T1) of the 13C nuclei
of the species in the MEA reaction mixture and of the standard
10169
dx.doi.org/10.1021/jp503421x | J. Phys. Chem. B 2014, 118, 10167−10174

The Journal of Physical Chemistry B
Article
in the capillary, the following parameters were used: recycle
delay of 120 s (corresponding to 6 times the longest T1), pulse
angle of 90° (8.9 μs pulse width) and 512 scans.18
between the hydroxyl and nitrogen functions increases, the
electron-withdrawing effect weakens, leaving the amino nitro-
gen surrounded by greater electron density and thereby
strengthening the basicity (Figure 1). Brønsted and Lewis
bases are both synonyms of nucleophiles but, for the Brønsted
bases, the proton is the only possible electrophile, which is the
reason why they are considered to be a subcategory of the more
encompassing Lewis bases.20
A typical 13C NMR spectrum of the carbons containing
species, observed and quantified at the equilibrium, is reported
in Figure 3. 13C NMR spectra, including assignments, for all the
amines in the reaction mixtures are documented in the SI. The
assignment of each signal to the corresponding carbon was
performed by means of 2D NMR experiments.
In Figure 4, the pKb of each amine21 (SI for the background)
is plotted against the amount of carbamate (in % as calculated
by eq 4).
To calculate the area integrals, the 13C NMR spectra were
fitted and the area of each peak was related to that of the C-2#
of the CH3CN standard. The area of the signal corresponding
to the C-* carbon of the amine carbamate and that
corresponding to the C-1# carbon of the standard were not
taken into account for the calculation of the concentrations of
the species in solutions because their T1 values are longer than
the longest T1 (20 s) which was used for setting the recycle
delay. Since the concentration of the amine carbamate species
could be determined by using carbons other than carbonyl and
the area of each carbon could be related to one of the signals
from the standard (i.e., C-2#), the recycle delay was set to be 6
times 20s. The longest T1 value was that of the methyl carbon
of the standard, and this allowed us to apply the same recycle
delay to all the reaction mixtures under study, avoiding T1
measurements of 13C nuclei in each of them.18
The fast exchanging proton species (neutral/protonated
amines and carbonate/bicarbonate) appear with a common
peak in the 13C NMR spectra and only the sum of their
concentration can be obtained. Various methods could be
utilized to estimate the contribution of each of these species but
this was not necessary for the present work.
Figure 4. Amine basicity (pKb, 293.15 K)21 as a function of the
percentage (%) of carbamate at equilibrium.
The analyses of the quantitative 13C NMR spectra for the
different amines were consistent with a decrease of the amount
For linear primary amines, the amount of carbamate at
equilibrium decreased with increasing basicity (lower pKb,
stronger bases). However, DEA and AMP did not fit into any
apparent correlation in this plot.
−
of carbonate species (HCO3 /CO32−) at increasing amount of
carbamate in solution (SI). Furthermore, in the reaction
mixtures of the amines with lower pKb (stronger bases), the
The carbamate forming reaction is the reaction of an amine
(Lewis base) with the electrophilic center (Lewis acid) of the
−
2−
carbons corresponding to HCO3 /CO3
species were
−
resonating at a higher chemical shift value which corresponds
HCO3 anion. The data presented in Figure 4 shows a
2−
−
to an higher ratio of CO3
(carbonate) to HCO3
discrepancy between the basicity, a function of the molecule’s
chemical structure, and the reactivity to form carbamate.
Indeed, the weaker bases, MEA and 1A2P, with the hydroxyl
function in the β position with respect to the nitrogen, would
be expected to have lower electron density on the nitrogen and,
consequently, relatively lower tendency to react, but we
observed the opposite.
(bicarbonate)19(SI). However, since the aim of the present
work is the study of the amine reactivity toward carbamate
formation, we will focus the discussion of the 13C NMR results
on the carbamate species only.
The amount of carbamate in the equilibrated reaction
mixtures was expressed in percentage with respect to the sum
of the concentrations of all the species detected in the 13C
NMR spectra, as shown in eq 4:
The basicity and expected electron density on the nitrogen of
DEA and AMP also did not reflect the predicted tendency to
form carbamate, but this behavior might be attributed to the
substitution effects and steric hindrance which reduce the
ability of the nitrogen to interact with the electrophilic center of
%carbamate = ([RNHCOO−]·100)
−1
× [RNHCOO−] + [RNH /RNH+] + [HCO−/CO2−]
(
)
2
3
3
3
−
the HCO3 anion. Similar findings have been reported by
(4)
Conway et al., who have related the protonation constants of
various amines with the kinetic and equilibrium constants for
the reaction of amine and CO2(aq) to carbamic acid/
carbamate.6,9b For the linear amines, a quite linear relationship
was observed and the deviation from that trend was attributed
to steric hindrance and substitution effects. However, ammonia,
which does not have any steric hindrance, also showed a
deviation; this was tentatively explained as being due to
different solvation properties.6
The error in the calculation of the % of carbamate was
estimated to be 0.76 percentage points which corresponded
to the standard deviation between the % of carbamate obtained
from three equilibrium experiments performed on MEA.
3. RESULTS AND DISCUSSION
The effect of molecular structure on base strength for the
current amines is related to inductive effects operating through
single bonds. Indeed, the hydroxyl function exerts an electron-
withdrawing inductive effect through bonds, whereas the alkyl
groups induce an electron-donating effect.11a,20 As the distance
However, pKb is a measure of proton accepting power of a
Brønsted base (e.g., the amine) in water. The above situation
indicates that not all underlying factors influencing the
10170
dx.doi.org/10.1021/jp503421x | J. Phys. Chem. B 2014, 118, 10167−10174

The Journal of Physical Chemistry B
Article
reactivity of amine to form carbamate may be reflected when
setting up a relationship, as shown in Figure 4.
nitrogen to attack an electrophile (the carbon of bicarbonate in
this study).
The 15N NMR chemical shift is a measure for the relative
electron density present on a N nucleus in respect to molecular
structure and medium effects. In Figure 5, the 15N chemical
Other factors that could influence the 15N chemical shifts,
such that at increasing basicity there are shifts to higher ppm
values (lower electron density), can be attributed to the
interactions of the nitrogen lone-pair with hydrogen of donor
groups, such as the water solvent and hydroxyl groups in other
amine molecules or in its own molecule.15a
Water is a protic solvent characterized by a relatively strong
polarity, as evidenced by the dielectric constant (ε = 78.4) and
the molecular dipolar moment (μ= 1.8). Moreover, it plays an
important role as hydrogen bond donor, since it exhibits a
hydrogen bond donor acidity (α = 1.17) higher than a
hydrogen bond acceptor basicity (β = 0.47).23 The hydrogen of
water can form hydrogen bonds with the unshared electron pair
on nitrogen and the degree of this interaction depends on the
degree of the lone pair delocalization. Hydrogen bonds to a
protic solvent should strengthen with increasing basicity and
the transition between intermolecular effects and chemical
reactions may not be clear.24 This is particularly true for proton
exchange reactions, such as amine protonation (RNH2 + H2O
⇆ RNH3+ + OH−). It is impossible to distinguish in the NMR
spectra the species exchanging a proton with water because the
proton transfer is faster than the NMR time scale at 298.15 K
and, as expected, only a single 15N NMR signal is observed for
both the solvated protonated and solvated free amine group in
the amine molecules. Hence, at the same concentration and
temperature, the 15N nucleus of the aqueous amines at
increased base strength will resonate at higher chemical shift
values (which depend on the relative amount of the solvated
free and solvated protonated amines) due to the interactions of
the water hydrogen with the amino nitrogen, in terms of
hydrogen bonds and/or protonation. An experimental con-
firmation of this expected 15N chemical shift trend is given by
further 15N NMR experiments performed on MEA and 3A1P.
Specifically, the chemical shift values of the 15N nuclei of the
amines, the fully protonated amines (amineH+) and the amine/
amineH+ (1:1) mixture in aqueous solutions at the same
concentration (2M) and temperature (298.15 K) were
recorded and compared (Table 1). The increased 15N ppm
values at increased protonation (i.e., MEA 18.38 ppm, MEA/
MEAH+ 24.34 ppm, MEAH+ 29.01 ppm and 3A1P 24.26 ppm,
3A1P/3A1PH+ 29.42 ppm, 3A1PH+ 33.66 ppm) are consistent
with the expected increase of the chemical shifts with increasing
interactions of the lone pair electrons of the amino nitrogen
with water.
The hydroxyl functionality on the structure of the amine
molecules also has an effect on the 15N chemical shifts in terms
of inter- and intramolecular hydrogen bonds. The latter were
computed by Yamada et al., who combined 13C NMR data and
computations to investigate secondary alkanolamines with
varied alkyl and alcohol chain length.7 Their analyses indicated
the likeliness of intramolecular H-bonds in alkanolamines if
allowed by the amine’s structure, i.e., if the molecular structure
is such that the −OH group can be aligned toward the N′s
electron lone pair.
In our experiments, the first evidence of such possible
interactions was observed for the ISOB molecule which lacks
the −OH functional group. At 2 M concentration, the nitrogen
resonated at a chemical shift value lower (higher electron
density) than the other primary amines of comparable basicity
and reactivity (i.e., 4A1B and 5A1P which have an −OH group
in the structure). Similarly, further evidence of such inter- and
Figure 5. 15N chemical shift (δ) of amines (2 M, 298.15 K) as a
function of the percentage (%) of carbamate formed in reaction
mixtures. Note that no bicarbonate was added when the 15N
measurements were done.
shift values of the amines at 2 M concentration (before
bicarbonate is added) are reported as a function of the % of
carbamate found at equilibrium (after the reaction of the
amines with bicarbonate).
Figure 5 shows a fairly good linear trend for all the data in
contrast to Figure 4. It appears that as the electron density
increased on the nitrogen (reflected in decreasing ppm values),
the amount of carbamate formed at equilibrium (after
bicarbonate addition) increased.
There was a clear linear relationship between the ability of
MEA, DEA, and AMP to form carbamate and their 15N
chemical shift values, a trend that was consistent with their
chemical structures. Indeed, the lower electron density on the
nitrogen of DEA as compared to MEA can be attributed to the
presence of two hydroxyl groups in the β-position relative to
the nitrogen, as compared to MEA’s one hydroxyl group. Even
though AMP has two methyl groups located α to the nitrogen,
the electron density on the nitrogen is relatively low in
comparison to the other amines. Chakraborty et al. showed that
the interaction between the nitrogen lone pair and the methyl
group orbitals can lead to significant changes in the donor
properties of the amino species, resulting from a higher and
more delocalized HOMO (Highest Occupied Molecular
Orbital) which leads to a lower charge on the donor nitrogen
site.22
For the unhindered primary amines (1A2P, MEA, 3A1P,
4A1B, 5A1P, and ISOB), the 15N chemical shift trend and,
consequently, the electron density on the nitrogen do not
appear to reflect their structure. With shorter distances between
the −OH and the −NH2 functional groups (decreased
basicity), the 15N nuclei should be more deshielded but, in
the case of these primary amines, the opposite is observed: the
15N nuclei of the weaker bases, such as MEA and 1A2P, have an
increased electron density (lower ppm values), resulting in an
increased relative availability of the unshared electrons on the
10171
dx.doi.org/10.1021/jp503421x | J. Phys. Chem. B 2014, 118, 10167−10174

The Journal of Physical Chemistry B
Article
intramolecular bonding effects was given by the 15N NMR
experiments performed on pure amines without any water
dilution (Table 1).
the amines to be protonated in this solvent is reduced (the
dissociation constants will be different). This finding is in
accordance with the reported decrease in amine base strength
with decreasing solvent dielectric constants.10
Carbamate formation in amine-H2O−DME mixtures was not
studied because sodium bicarbonate is unsoluble in such
solutions. Moreover, further attempts with other solvents were
not carried out, as such investigations were beyond the scope of
the current work. However, since the 15N chemical shifts reflect
the tendency to form carbamate, it may be assumed that the
equilibrium for the carbamate formation in H2O/DME would
be shifted toward higher amounts of carbamate than that
observed in 100% water.
Our findings that the tendency for carbamate formation is
well described by the 15N chemical shift values for these
different molecules advances the understanding of the
chemistry involved in aqueous amine solutions. Specifically,
the solvent effect has been identified to be an additional factor
influencing the reactivity of the amines.
As expected, in neat preparations, the nitrogen atom of each
molecule resonates at a chemical shift lower than the
corresponding aqueous amines at 2 M concentration because
of the lack of water interactions. However, the stronger pure
bases would be expected to have 15N chemical shift values
lower than the weaker pure ones. But, a trend similar to the
diluted aqueous amines was observed, suggesting the presence
of similar interactions which involve −OH and −NH2
functional groups of the amine structure. Comparisons of the
chemical shift differences (Δδ) between the pure and diluted
amines indicated that the linear primary alkanolamines show
similar Δδs (for MEA it is 1.54 ppm, for 3A1P it is 1.37 ppm,
and for 4A1B it is 1.33 ppm), whereas for 1A2P, the value is
0.45 ppm and for ISOB, 3.45 ppm. The smaller Δδ for 1A2P as
compared to the other molecules can probably be attributed to
the particular position of the hydroxyl group. Compared to the
linear primary alkanolamines, 1A2P could be more restricted to
movements (more rigid) so that the chemical shifts are similar
either with or without water. For ISOB, the large Δδ observed
may be ascribed to the lack of inter- and intramolecular
hydrogen bondings of the nitrogen atom to the hydroxyl group
on the amine.
4. CONCLUSIONS
In this study, 13C and 15N NMR spectroscopy was used to
investigate the relationship between the chemical properties of
amines and their tendency to form amine carbamates. For all
the amines under study, the 15N chemical shift values reflected
the observed reactivity for forming carbamates very well.
Brønsted and Lewis bases are both synonyms of nucleophiles
but, for the Brønsted bases, the proton is the only possible
electrophile. The pKb is indeed a measure of the proton
accepting strength of a Brønsted base (e.g., the amine) in water.
The 15N NMR chemical shift is a measure of the relative
electron density present on the N atom in respect to molecular
structure and medium effects. In the current study, 15N NMR
chemical shift data are able to fit all our carbamate formation
data (reaction of an amine Lewis base, the nucleophile, with the
Our 15N NMR data for linear primary alkanolamines would
suggest that both solvent interactions and inter/intra molecular
hydrogen bonds between functional groups on the amine
structure could influence the carbamate formation reaction, but
the molecular structure of the alkylamine ISOB provided
insight into the main component affecting the reactivity of the
studied unhindered primary amines. ISOB has a base strength
and reactivity similar to 4A1B and 5A1P, but differs by the
absence of hydrogen bonds between the nitrogen free electron
pair and the hydroxyl function on the structure. Therefore, the
main factor influencing the reactivity of these primary
unhindered amines can be considered to be the interaction of
the nitrogen electron lone-pair with water. The stronger the
base, the more interactions with water occur, leading to a more
solvation (which also involves the protonated amines) and,
consequently, to the need of desolvation before nucleophilic
attack for carbamate formation.25
Moreover, the hydroxyl function in the β-position to the
nitrogen atom of the current unhindered primary amines (MEA
and 1A2P) showed to increase the amine reactivity due to the
electron-withdrawal property which reduced the basicity and,
consequently, led to a relative weaker solvation. In the case of
DEA and AMP, substitution effects and steric hindrance should
also be considered. The structures of DEA and AMP differ from
the other studied amines, but intra- and intermolecular H-bond
networks cannot be excluded.
To further investigate the role played by the solvent on the
availability of the lone pair electrons on the nitrogen atom, we
performed 15N NMR experiments on MEA and AMP in solvent
blends containing both polar-protic and -aprotic components.
Specifically, MEA and AMP solutions at 2 M concentration
were prepared in water (H2O)/dimethoxyethane (DME) at 1:1
ratio and the 15N chemical shift values were compared to those
obtained for the same amines in 100% water (Table 1). DME
has a dielectric constant (ε = 7.2) lower than that of water (ε =
78.4) and the chemical shift values of the amines in DME-H2O
solvent are lower than in water alone. Thus, the electron
density on the nitrogen is increased because the tendency of
−
Lewis acid center, the electrophile, of the HCO3 ) into a linear
relationship, in contrast to a pKb based relationship.
We demonstrated that the amount of amine carbamate
formed by MEA and related unhindered primary amines
decreased at increasing basicity due to the water solvent effect,
which influenced the reactivity of the amine group. Under
equivalent reaction conditions, the stronger unhindered bases
were shown to have less availability of the N lone pair electrons
to attack an electrophilic carbon for carbamate formation. This
was attributed to their higher tendency to interact with the
solvent. Such solvent effects have thus far been underestimated
in the field of chemical absorption of CO2. Concerning DEA
and AMP, substitution and steric hindrance make their
structures to some degree quite different from the other
amines. However, their reactivity was well reflected by the 15N
NMR chemical shifts, and was therefore dependent on the
availability of lone pair electrons on this nucleus, but the main
factors influencing this have not been identified in this study.
In the field of chemical absorption of CO2 by amine
absorbents, these findings represent a step toward under-
standing the underlying dynamics of reactivity and, based on
15N chemical shift values, allow us to estimate the amine
activity.
10172
dx.doi.org/10.1021/jp503421x | J. Phys. Chem. B 2014, 118, 10167−10174

The Journal of Physical Chemistry B
Article
(7) Yamada, H.; Matsuzaki, Y.; Chowdhury, F.; Higashii, T.
Computational investigation of carbon dioxide absorption in alkanol-
amine solutions. J. Mol. Model. 2013, 19 (10), 4147−4153.
(8) Puxty, G.; Rowland, R.; Allport, A.; Yang, Q.; Bown, M.; Burns,
R.; Maeder, M.; Attalla, M. Carbon Dioxide Postcombustion Capture:
A Novel Screening Study of the Carbon Dioxide Absorption
Performance of 76 Amines. Environ. Sci. Technol. 2009, 43 (16),
6427−6433.
(9) (a) Versteeg, G. F.; van Swaaij, W. P. M. On the kinetics between
CO2 and alkanolamines both in aqueous and non-aqueous solutions
I. Primary and secondary amines. Chem. Eng. Sci. 1988, 43 (3), 573−
585. (b) Conway, W.; Wang, X.; Fernandes, D.; Burns, R.; Lawrance,
G.; Puxty, G.; Maeder, M. Toward Rational Design of Amine Solutions
for PCC Applications: The Kinetics of the Reaction of CO2(aq) with
Cyclic and Secondary Amines in Aqueous Solution. Environ. Sci.
Technol. 2012, 46 (13), 7422−7429.
ASSOCIATED CONTENT
■
S
* Supporting Information
pKb background of the amines, 15N- and 13C- NMR spectra,
and a graph on the 13C NMR spectra analysis. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Tel.: +47-35575193; fax: +47-35575001; e-mail: Klaus.J.
Jens@hit.no.
Notes
The authors declare no competing financial interest.
(10) Hamborg, E. S.; van Aken, C.; Versteeg, G. F. The effect of
aqueous organic solvents on the dissociation constants and
thermodynamic properties of alkanolamines. Fluid Phase Equilib.
2010, 291 (1), 32−39.
(11) (a) Mergler, Y.; Gurp, R.R.v.; Brasser, P.; Koning, M.d.;
Goetheer, E. Solvents for CO2 capture. Structure-activity relationships
combined with vapour−liquid−equilibrium measurements. Energy
Procedia 2011, 4, 259−266. (b) Ciftja, A. F.; Hartono, A.; Svendsen,
H. F. Carbamate Formation in Aqueous−diamine−CO2 Systems.
Energy Procedia 2013, 37 (0), 1605−1612.
ACKNOWLEDGMENTS
■
Financial assistance, a scholarship (to C.P.) provided by the
Research Council of Norway (CLIMIT grant nr. 199890), and
support from the SINTEF NMR lab, including staff are
gratefully acknowledged.
ABBREVIATIONS
■
MEA, 2-amino-1-ethanol or ethanolamine; 3A1P, 3-amino-1-
propanol; 4A1B, 4-amino-1-butanol; 5A1P, 5-amino-1-penta-
nol; ISOB, isobutylamine; 1A2P, 1-amino-2-propanol; DEA,
2,2′-iminodiethanol or diethanolamine; AMP, 2-amino-2-
methyl-1-propanol; DME, dimethoxyethane; STP conditions,
(12) Han, B.; Sun, Y.; Fan, M.; Cheng, H. On the CO2 Capture in
Water-Free Monoethanolamine Solution: An ab Initio Molecular
Dynamics Study. J. Phys. Chem. B 2013, 117 (19), 5971−5977.
(13) Arstad, B.; Blom, R.; Swang, O. CO2 Absorption in Aqueous
Solutions of Alkanolamines: Mechanistic Insight from Quantum
Chemical Calculations. J. Phys. Chem. A 2007, 111 (7), 1222−1228.
(14) da Silva, E. F. Theoretical study of the equilibrium constants for
solvents for CO2 capture. Energy Procedia 2011, 4 (0), 164−170.
(15) (a) Duthaler, R. O.; Roberts, J. D. Nitrogen-15 nuclear magnetic
resonance spectroscopy. Solvent effects on the 15N chemical shifts of
saturated amines and their hydrochlorides. J. Magn. Reson. (1969)
−
Standard Temperature Pressure conditions; HCO3 , bicarbon-
ate; CO32−, carbonate; CO2, carbon dioxide; CH3CN,
acetonitrile; AmineH+, protonated amine; H2O, water; NMR,
Nuclear Magnetic Resonance; OH, hydroxyl functional
group; NH2, amino functional group in primary amines;
NH, amino functional group in secondary amines; 
NCOO−, amino functional group in amine carbamates; %,
percent; δ, chemical shift; Δδ, chemical shift differences; T1,
longitudinal relaxation time constant; s, seconds; h, hour; μs,
microseconds; K, Kelvin (unit of temperature); ppm, parts per
million
1979, 34 (1), 129−139. (b) von Philipsborn, W.; Muller, R. 15N-NMR
̈
SpectroscopyNew Methods and Applications [New Analytical
Methods (28)]. Angew. Chem. Int. Ed. Eng. 1986, 25 (5), 383−413.
(16) Yoon, S. J.; Lee, H. Substituents Effect in Amine−CO2
Interaction Investigated by NMR and IR Spectroscopies. Chem. Lett.
2003, 32 (4), 344−5.
(17) Harris, R. K.; Becker, E. D.; Cabral de Menezes, S. M.;
Goodfellow, R.; Granger, P. NMR nomenclature: Nuclear spin
properties and conventions for chemical shifts. IUPAC Recommen-
dations 2001. International Union of Pure and Applied Chemistry.
Physical Chemistry Division. Commission on Molecular Structure and
Spectroscopy. Magn. Reson. Chem. 2002, 40 (7), 489−505.
(18) Perinu, C.; Arstad, B.; Bouzga, A. M.; Svendsen, J. A.; Jens, K.-J.
NMR-Based Carbamate Decomposition Constants of Linear Primary
Alkanolamines for CO2 Capture. J. Ind. Eng. Chem. Res. 2014,
submitted for publication.
(19) (a) Jakobsen, J. P.; Krane, J.; Svendsen, H. F. Liquid-Phase
Composition Determination in CO2H2O−Alkanolamine Systems:
An NMR Study. Ind. Eng. Chem. Res. 2005, 44 (26), 9894−9903.
(b) Perinu, C.; Arstad, B.; Jens, K.-J. NMR spectroscopy applied to
amine−CO2−H2O systems relevant for post-combustion CO2 capture:
A review. Int. J. Greenhouse Gas Control 2014, 20, 230−243.
(20) Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic
Chemistry; University Science Books: Sausalito, CA, 2006, 284−285.
(21) (a) Perrin, D. D. Dissociation Constants of Organic Bases in
Aqueous Solution; Butterworths, Scientific Publications: London, 1965
(b) Perrin, D. D. Dissociation Constants of Organic Bases in Aqueous
Solution; Butterworths, Scientific Publications: London, 1972.
(22) Chakraborty, A. K.; Bischoff, K. B.; Astarita, G.; Damewood, J.
R. Molecular orbital approach to substituent effects in amine−CO2
interactions. J. Am. Chem. Soc. 1988, 110 (21), 6947−6954.
REFERENCES
■
(1) Rochelle, G. T. Amine Scrubbing for CO2 Capture. Science 2009,
325 (5948), 1652−1654.
(2) Reynolds, A. J.; Verheyen, T. V.; Adeloju, S. B.; Meuleman, E.;
Feron, P. Towards Commercial Scale Postcombustion Capture of CO2
with Monoethanolamine Solvent: Key Considerations for Solvent
Management and Environmental Impacts. Environ. Sci. Technol. 2012,
46 (7), 3643−3654.
(3) Xie, H.-B.; Johnson, J. K.; Perry, R. J.; Genovese, S.; Wood, B. R.
A Computational Study of the Heats of Reaction of Substituted
Monoethanolamine with CO2. J. Phys. Chem. A 2010, 115 (3), 342−
350.
(4) Fernandes, D.; Conway, W.; Burns, R.; Lawrance, G.; Maeder,
M.; Puxty, G. Investigations of primary and secondary amine
1
carbamate stability by H NMR spectroscopy for post combustion
capture of carbon dioxide. J. Chem. Thermodyn. 2012, 54 (0), 183−
191.
(5) Sartori, G.; Savage, D. Sterically Hindered Amines for CO2
Removal from Gases. Ind. Eng. Chem. Fundam. 1983, 22, 239−249.
(6) Conway, W.; Wang, X.; Fernandes, D.; Burns, R.; Lawrance, G.;
Puxty, G.; Maeder, M. Toward the Understanding of Chemical
Absorption Processes for Post-Combustion Capture of Carbon
Dioxide: Electronic and Steric Considerations from the Kinetics of
Reactions of CO2(aq) with Sterically hindered Amines. Environ. Sci.
Technol. 2013, 47 (2), 1163−1169.
10173
dx.doi.org/10.1021/jp503421x | J. Phys. Chem. B 2014, 118, 10167−10174

The Journal of Physical Chemistry B
Article
(23) Carrol, F. A. Perspectives on Structure and Mechanism in Organic
Chemistry; Brooks/Cole Publishing: Monterey, CA, 1998, 329−332.
(24) Martin, G. T.; Martin, M. L.; Gouensnard, J. P. 15N NMR
Spectroscopy; Springer-Verlag: Berlin/Heidelberg/New York, 1981, 18,
54−74.
(25) Jencks, W. P., Effects of Solvation on Nucleophilic Reactivity in
Hydroxylic Solvents. In Nucleophilicity; American Chemical Society:
Washington, DC, 1987; Vol. 215, pp 155−167.
10174
dx.doi.org/10.1021/jp503421x | J. Phys. Chem. B 2014, 118, 10167−10174