Subscriber access provided by HENKEL KGAA
Molecular orbital approach to substituent effects in amine-CO2 interactions
A. K. Chakraborty, K. B. Bischoff, G. Astarita, and J. R. Damewood
J. Am. Chem. Soc., 1988, 110 (21), 6947-6954 • DOI: 10.1021/ja00229a003
Downloaded from http://pubs.acs.org on December 4, 2008
More About This Article
The permalink http://dx.doi.org/10.1021/ja00229a003 provides access to:
•
•
Links to articles and content related to this article
Copyright permission to reproduce figures and/or text from this article
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
J . Am. Chem. SOC.1988, 110, 6947-6954
tivation due to accumulation of P on the catalyst. For example,
P from decomposed PH3diffuses in and out of Pt( 1 1 1) above 300
K and is very difficult to remove by oxidation, espeically a t high
coverages7 The key to these oxidative processes is the removal
of P by PO, formation and desorption before atomic P forms.
In a recent study by Smentkowski et al.1° catalytic decomposition of D M M P was observed on Mo(l10) above 898 K in a 2:l
0,-DMMP molecular beam without accumulation of P or C on
the surface. The Mo(l10) surface formed a stable oxide layer
that catalyzed D M M P decomposition but prevented accumulation
and bulk diffusion of P. Our results for D M M P decomposition
on a-Fe2039 show that it is active for D M M P decomposition but
deactivates as phosphate forms.
Unlike Mo, Pt does not readily form a stable bulk oxide. Thus,
removal of P can occur only when gas-phase oxygen is present
these PO,
to drive the formation of PO,. In the absence of 02,
species are stable only below 500 K. The dominant PO, species
observed in this study is PO. In the absence of 0, it decomposes
with an activation barrier of 35.9 kcal/mol, about 23 kcal/mol
lower than its desorption barrier from Pt." Thus removal of PO
in a catalytic process of Pt will require high temperature to overcome the desorption barrier and high pressures of 0, to maintain
P-O bonds. Our results suggest that, in the absence of 02, M M P
D
decomposition on Pt( 11 1) above 600 K will lead to phosphorus
6947
accumulation and catalytic deactivation.
5. Summary
(1) At 100 K on Pt(l1 l), D M M P adsorbs molecularly through
the oxygen lone pairs of the -p--Ogroup. This structure is stable
to 300 K.
(2) TPD after a multilayer exposure of D M M P at 100 K gives
multilayer desorption (220 K) and 0.09 M L of D M M P decomposition. The products are 0.27 M L of C O (437, 505, 586 K),
0.79 M L of H2 (340,448, 504 K), surface P, and a small amount
of surface C.
(3) TPSSIMS results indicate that, between 300 and 500 K,
D M M P decomposes in steps to PO,. Near 300 K PO-C bonds
cleave, with some P-OC bond cleavage at high-DMMP coverages.
At higher temperatures, P-C bonds cleave, leaving at least two
PO, species on the surface.
(4) The dominant PO, species is PO, which is characterized
by a 1255-cm-I u(P=O) loss in HREELS and an intense PO'
SSIMS signal. It is stable to a t least 500 K.
Acknowledgment. This research was supported in part by the
U S . Army Research Office. We gratefully acknowledge helpful
discussions with Dr. Rama Hegde.
Registry No. Pt, 7440-06-4; DMMP, 756-79-6.
Molecular Orbital Approach to Substituent Effects in
Amine-C02 Interactions
A. K. Chakraborty, K. B. Bischoff,* G . Astarita, and J. R. Damewood, Jr.
Contribution from the Departments of Chemical Engineering and Chemistry and The Center for
Catalytic Science and Technology, Uniuersity of Delaware, Newark, Delaware I971 6.
Received April 3, 1987
Abstract: The effects of substituents at the 0-carbon atom on the donor properties of primary amines and amino alcohols
have been studied. Such substituted amino species have important applications in industrially relevant gas-separation processes.
Qualitative molecular orbital arguments, along with detailed calculations at the MNDO level of theory, show that upon methyl
substitution at the a-carbon atom the interactions of the methyl group orbitals with the nitrogen lone-pair orbital lead to subtle
but significant changes in the donor properties of the amino species. Infrared spectroscopic data supporting the calculations
are also described. The implications of changes in the donor properties of the amino species on its reactions with acidic gases,
such as CO1, have been considered. These studies serve to elucidate previous phenomenological results regarding the macroscopic
kinetic and equilibrium behavior of substituted amino species in C 0 2 absorption applications.
1. Introduction
The removal of acidic gases such as CO,,
H2S, and SO2 from
gas streams is widespread in the chemical processing industries.'
This separation is commonly accomplished by absorbing the acidic
gas into a chemically reactive liquid. Carrying out such processes
efficiently is of crucial importance from several viewpoints, and
a detailed understanding of these reacting, two-phase systems is
therefore necessary.
The absorption of an acidic gas into a chemically reacting liquid
is a mass-transfer process accompanied by a Lewis acid-base
reaction. There are therefore two aspects of the problem that need
consideration. On one hand, the facilitated-transport problem
of mass transfer with chemical reaction needs to be understood
within the framework of macroscopic phenomenological models.
On the other hand, a comprehensive understanding of the system
is possible only if the fundamental molecular chemistry and its
macroscopic manifestations are clearly elucidated. Such an un(1)
Charpentier, J. C. Chem. Eng. Res. Des. 1982, 60, 131.
0002-7863/88/1510-6947$01.50/0
derstanding is crucial if one is to attain the goal of being able to
"tailor-make" molecules that could carry out chemical absorption
based separations in a very specific manner.
While the literature (particularly in engineering science) is rich
with work on the facilitated-transport p r ~ b l e m , ~ - ~
relatively little
emphasis has been placed on a fundamental understanding of the
molecular chemistry of these systems. In this paper, the reactions
of a model acidic gas (CO,) with various amines and amino
alcohols are investigated. In particular, the effects of substituents
on the Lewis acid-base interaction is investigated within the
theoretical framework of perturbation molecular orbital theory6
(2) Astarita, G.; Bisio, A.; Savage, D. W. Gas Treating Wirh Chemical
Solvents; Wiley-Interscience: New York, 1983.
(3) Danckwerts, P. V. Gas-Liquid Reactions; McGraw-Hill: New York,
1970.
(4)
(5)
(6)
Astarita, G.; Savage, D. W. Chem. Eng. Sci. 1980, 35, 659.
Danckwerts, P. V.; Sharma, M. M. Chem. Eng. (N.Y.) 1966, 244.
Fukui, K. In Molecular Orbitals in Physics, Chemistry and Biology;
Lowdin, P. 0.. Pullman, B., Eds.; Academic: New York, 1964.
0 1988 American Chemical Society
Chakraborty et al.
6948 J . Am. Chem. SOC., 110, No. 21. 1988
Vol.
Table 1. Examples of the Primary Amines Introduced by Sartori and
SavageI2 for Acid Gas-Separation Applications
structure
name
~~
CH3
I
I
2-amino-2-methyl- 1-propanol (AMP)
HOHzC-C-NH2
CH3
CHI
HOOC
I
I
2-amino-2-methyipropionic acid
-C - NH2
CH3
HJC-C-CH~
I
NH2
and detailed molecular orbital calculations using the M N D O
method.'
Experiments that support the results obtained by
molecular orbital theory are also described. The understanding
of the effects of substituents on the donor properties of amino
species resulting from these studies is then related to previous
macroscopic kinetic and equilibrium observations.8-12
The reaction of a Lewis acid such as C02 with a Lewis base
such as a n amine or a n amino alcohol (wherein the donor atom
is nitrogen) represents a classical donoracceptor interaction. The
donor orbital in this case is nonbonding, while the acceptor orbital
is antibonding. On the basis of their own macroscopic kinetic and
equilibrium data, data in the literature, and the work of Astarita
and co-workersI3 and Danckwerts,l* Chakraborty et
have
postulated the reaction pathway (eq 1-4) for the reactions of C02
COZ + RNH2 = [COZ:RNH2] = R N H C O O -
+ H+
+ OH- = HC03- + R N H 2
RNH2 + H" = RNH3"
H20 = H+ + OH-
[C02:RNH2]
(1)
(2)
(3)
(4)
with primary amines in aqueous solutions. In the reaction pathway
shown above, R represents the substituent attached to the amino
nitrogen. This reaction pathway involves a n amine-COz interaction to form a 1:l acid-base adduct, which may then undergo
two different transformations. If the N-C bond in the adduct
is strong, there will be a concomitant depopulation of the N-H
bonding (IIN~J
orbitals. If sufficiently strong, the depopulation
of these orbitals may lead to the lass of a proton and the formation
of the carbamate species. In aqueous solution, the acid-base
adduct may also react with OH- (or water) and a base displacement reaction may take place resulting in the formation of
the bicarbonate species. In addition, amine protonation and water
dissociation occur in solution as represented by steps 3 and 4. If
the step leading to the formation of the carbamate species is
neglected, the proposed pathway corresponds to a classical homogeneous catalysis mechanism for the formation of the bicarbonate species.
In 1983 Sartori and SavageI2 presented a significant advance
in acid gas-separation technology by introducing the use of a new
class of amines, which proved to have higher rates and equilibrium
(7) Dewar, M. J. S.;
Thiel, W. J . Am. Chem. Soc. 1977, 99, 4899.
(8) Laddha, S. S.;Danckwerts, P. V. Chem. Eng. Sci. 1982, 36, 475.
(9) Alvarez Fuster, C.; Midoux, N.; Laurent, A.; Charpentier, J. C. Chem.
Eng. Sci. 1980, 33, 1717.
(10) Chakraborty, A. K.; Astarita, G.; Bischoff, K. B. Chem. Eng. Sci.
1986, 41, 997.
(1 1 ) Chakraborty, A. K.; Astarita, G.; Bischoff, K. B., to be submitted for
publication in AIChE J .
(12) Sartori, G.; Savage, D. W. Ind. Eng. Chem. Fundam. 1983,22,239.
(13) Astarita, G.;Savage, D. W.; Longo, J. M. Chem. Eng. Sci. 1981,36,
---.
511 1
(14) Danckwerts, P. V. Chem. Eng. Sci. 1979, 34,443.
capacities of absorption than the amino species used previously.
Primary amines wherein the amino nitrogen is attached to a
tertiary carbon atom are members of this class. Some examples
are listed in Table I. Many of the primary amines introduced
by Sartori and SavageI2 for acid gas-separation applications (and
certainly those listed in Table I) have methyl substituents attached
to the a-carbon atom. Sartori and Savage attributed the favorable
behavior of these amines to the relative instability of the carbamate
species. Chakraborty et
have reported I3C N M R data that
show that the carbamate of 2-amino-2-methyl- 1-propanol is
formed to a much lesser extent than the carbamate of the corresponding unsubstituted amine (ethanolamine). Sartori and
SavageI2 ascribed the instability of the carbamate species to steric
constraints imposed by the alkyl groups attached to the amino
group. However, both Sartori and Savagei2 and Chakraborty and
co-workersIo*lIprovide only intuitive explanations for the cause
of this lower carbamate stability and their other kinetic and
equilibrium observations.
In this work, a further examination of the effects of substituents
on the carbon atom adjacent to the amino group is described. In
particular, the manifestations of changes in the electronic and
structural characteristics of these amines upon such substitutions
in the acid-base reactions under consideration are Avestigated
in detail.
2. Effect of Substituents
It has been recognized for a long time that acids and bases
cannot be completely characterized by a single parameter such
as the pK,. One successful attempt at rationalizing the vast body
of experimental data available in the literature is the hard and
soft acid-base (HSAB) principle developed by Pearson and coworkers.I5-l8 The HSAB principle characterizes acids and bases
as being either hard or soft by using properties such as charge,
size, extent of electron delocalization and distortability, etc.
Pearson and co-workers rationalized the available data by formulating the hypothesis that like species prefer to interact with
each other. This generalization applies to both the equilibrium
and the rates of the reactions under consideration.
These empirical generalizations were put on a more quantitative
basis by KlopmanI9 within the framework of perturbation molecular orbital theory.20 Klopman21 developed the following
expression for the energy of interaction of two species R and S,
interacting a t the sites r and s (eq 5) where qi is the charge a t
the ith interacting site; cj is the coefficient of the atomic orbital,
i, in the molecular orbital, j; P is the resonance integral; Ei corresponds to the energy of the molecular orbital, i; e is the dielectric
'
constant of the medium; I is the coulomb repulsion term; Asolvn
is the energy associated with desolvation; and AE represents the
energy associated with the partial formation of a bond (in a given
medium) between R and S .
Since eq 5 was developed within the framework of P M O theory,
it is rigorously applicable to the early part of the reaction coordinate. The first term (the 'charge-control" term) represents the
coulombic interaction between the donor and the acceptor sites.
The second term (the "orbital-control" term) represents the interactions of the molecular orbitals of the interacting species. If
one considers only the interactions of the frontier orbitals then
E , and E,, correspond to the energies of the highest occupied
molecular orbital (HOMO) of the donor species and the lowest
unoccupied molecular orbital (LUMO) of the acceptor species,
respectively. Hence, for the interaction of a given acid (Le. with
( 1 5) Pearson, R. G. Hard andSofi Acids and Bases; Dowden, Hutchinson
and Ross Inc.: Stroudsburgh, PA, 1973.
ot
(16) Ho,T.L. Hard and S f Acids and Bases Principle in Organic
Chemistry; Academic: New York, 1977.
(17) Pearson, R. G. Sum. Prog. Chem. 1969, 5 , 1 .
(18) Pearson, R. G. J . Chem. Educ. 1987, 64, 561.
(19) Klopman, G. J . Am. Chem. SOC.1968, 90, 223.
(20) Klopman, G. Tetrahedron Lett. 1967, No. 12, 1103.
(21) Klopman, G. Theor. Chim. Acta 1967, 8, 165.
J. Am. Chem. Soc.. Vol. 1 1 0, No. 21, 198% 6949
Amine-C02 Interactions
a fixed value of E”), as long as E , < E,, the higher the energy
of the HOMO, the greater the importance of the orbital-control
term. Similarly, the lower the charge a t the donor site, the less
important the charge-control term. These and other arguments18J2
show that a donor species with a high-lying, more delocalized (or
diffuse) H O M O with a low charge a t the donor site corresponds
to a soft base in the Pearson sense. Although molecular orbital
arguments will be used frequently in this paper, for ease of reference, a donor with a higher and more delocalized H O M O and
a lower charge a t the donor site will therefore be referred to as
a “softer” base.
A series of papers authored by Parr and ~ o - w o r k e r s ~ ~ - ~ ~
pioneered the use of density functional theory% in chemistry. Within
the framework of density functional theory, Pearson and Parr27-29
have shown that a species may be characterized by its electronegativity (or potential), x (or p ) , and its hardness, 7. Within
the framework of density functional theory these quantities are
defined in finite difference form as shown in eq 6 and 7
x
= -p = ( I
7)
+ A)/2
= (I-A)/2
(6)
(7)
wherein I is the ionization potential and A is the electron affinity.
Pearson and Parr have shown that the lower the value of 7, the
softer the species. Since the value of 7 reflects the HOMOL U M O energy separation in a given species, this implies that the
smaller this energy separation, the softer the species. Pearson and
Parr’s work within the framework of density functional theory
also serves to place the H S A B concept on a more fundamental
and quantitative basis.
As mentioned earlier, many of the amino species that have
recently been found to be useful in acid gas-separation applications
have methyl substituents attached to the carbon atom adjacent
to the amino group. One of the goals of this paper is to examine
a t a molecular level the effects of substituents on the behavior
of amino species in industrially relevant gas-absorption applications. Therefore, it is appropriate to first examine the effects of
methyl substitution a t the a-carbon atom on the electronic environment a t the donor site, nitrogen. Changes in the electronic
environment a t the donor site will determine the donor properties
of the amino species and hence its interactions with Lewis acids.
Some noteworthy experimental studies on the effects of substituents on the basicities of amines (particularly with regard to proton
affinities) have been carried out by M ~ n s o n Brauman et al.,31
,~~
Aue and c o - w o r k e r ~ , Wooden and B e a ~ c h a m pand~Arnett
~~*~~
,~
and c o - w o r k e r ~ . ~ ~
In order to examine the effects of methyl substitution a t the
a carbon atom attention should be focussed on the possible interactions of the nonbonding nitrogen lone-pair orbital with the
orbitals of methyl group. Methyl groups possess two pairs of
degenerate orbitals of II-type symmetry,36 that are symmetry
compatible with the lone-pair orbital. One pair is occupied (IIMe),
(22) Pullman, A.; Brochen, P. Chem. Phys. Lett. 1975, 34, 7.
(23) Parr, R. G.; Donnelly, R. A.; Levy, M.; Palke, E. J . Chem. Phys.
1978, 68, 3801.
(24) Parr, R. G.; Bartolotti, L. J. J . Am. Chem. SOC.1982, 104, 3801.
(25) Parr, R.G.; Yang, W. J . Am. Chem. SOC.
1984, 106, 4049.
(26) Hohenberg, P.; Kohn, W. Phys. Reu. 1964, 136, B864.
(27) Parr, R. G.; Pearson, R. G. J . Am. Chem. SOC.1983, 105, 7512.
(28) Pearson, R. G. J . Am. Chem. SOC.1985, 107, 6801.
(29) Pearson, R. G . J . Am. Chem. SOC.1986, 108, 6109.
(30) Munson, M. S. B. J . Am. Chem. SOC.
1965, 87, 2332.
(31) Brauman, J. I.; Riveros, J. M.; Blair, L. K. J . Am. Chem. Soc. 1971,
93, 3914.
(32) Aue, D. H.; Webb, H. M.; Bowers, M. T. J . Am. Chem. SOC.
1976,
98, 311.
(33) Aue, D. H.; Webb, H. M.; Bowers, M. T. J . Am. Chem. SOC.1972,
94, 4726.
(34) Wooden, R. L.; Beauchamp, J. L. J . Am. Chem. Soc. 1978,100, 501.
(35) Arnett, E. M.; Jones, F. M.; Taagepera, M.; Henderson, W. G.;
Beauchamp, J. L.; Holtz, D.; Taft, R.W. J. Am. Chem. SOC.
1972, 94, 4724.
(36) Jorgensen, W. L.; Salem, L. The Organic Chemist’s Handbook of
Orbitals; Academic: New York, 1973.
TMe*
-,
\
\
\
\
\
/
/
/
/
/
I
/
/
/
H Me
w’
Figure 1. Energy diagram depicting the interactions of the nonbonding
nitrogen lone-pair orbital with the IIMc
and IIMc*
orbitals of the methyl
substituents. For simplicity the interaction with only one of each pair
of methyl group orbitals exhibiting II-type symmetry is shown.
and the other is unoccupied (IIM,*).One possible interaction is
that of the lone-pair orbital with the symmetry compatible
IIM,-filled orbitals of the methyl groups. Pellerute and B r a ~ m a n ~ ~
have considered the possibility of such interactions while investigating the effects of methyl substitution on the chemistry of
carbon acids. Pross and R a d ~ m ~ DeFrees et al.39 have also
and ~
considered the possibility of such interactions while examining
the effects of methyl substitution a t anionic centers. This interaction leads to an overall destabilization and thus would result
in a higher energy, more delocalized H O M O . There would,
however, be no significant redistribution of electron density as
a result of this interaction. The result of this interaction is consistent with the traditional view that the methyl group acts as an
electron d o n ~ r . ’ ~ ? is ~
It ~ worth noting, however, that, being a
four-electron interaction, it is inherently weak; furthermore, since
in the case under consideration here the methyl substituents are
one carbon atom removed, the effects of this interaction should
be further attenuated.
Another potentially important interaction is that between the
orbitals of
nitrogen lone-pair orbital and the antibonding IIMc*
the methyl substituents. Such a II back-bonding is a two-electron
interaction and hence should be an inherently stronger interaction
than the four-electron interaction considered above. However,
orbitals than
the lone-pair orbital lies closer in energy to the IIMe
to the IIMc*
orbitals, and this factor favors the interaction with
the filled IIMe
orbitals. The relative strength of the interactions
of the lone-pair orbital with the IIMcand IIMe*orbitals is determined by both these factors. Hehre and Popleu have carried
out a b initio calculations for amines with methyl substituents a t
the heteroatom, nitrogen, and have observed a back-donation from
the nitrogen to the methyl groups. Fujimoto et
have also
carried out calculations for amines with methyl substituents a t
the heteroatom. DeFrees et al.39 and Pross and R a d ~ m ~ ~
have
considered the interactions of the II-type methyl orbitals with those
of the parent molecule while discussing the effects of methyl
substitution a t anionic centers and on the chemistry of carbon
acids, respectively. The interaction of the nitrogen lone-pair orbital
(37) Pellerute, M. J.; Brauman, J. I. In Comprehenriue Carbanion Chemistry; Buncell, E., Durst, T., Eds.; Elsevier: Amsterdam, The Netherlands,
1980; Part A.
(38) Pross, A.; Radom, L. J. Am. Chem. SOC.1978, 100, 6512.
(39) DeFrees, D. J.; Bartness, J. E.; Kim, J. K.; McIver, R. T.; Hehre, W.
J. J . Am. Chem. SOC.1977, 99, 6451.
(40) Hehre, W. J.; Pople, J. A. J. Am. Chem. SOC.1970, 92, 2191.
(41) Fujimoto, H.; Yamasaki, T.; Hataue, I.; Koga, N. J . Phys. Chem.
1985, 89, 179.
Chakraborty et al.
6950 J . Am. Chem. SOC.,
Vol. 110, No. 21, 1988
with the IIMe*
antibonding orbitals of the methyl substituents has
several important implications. This interaction has a net stabilizing effect, but more importantly, it leads to the redistribution
of electron density. Hehre and Pople“ have observed that upon
methyl substitution a t the heteroatom this interaction leads to a
lower charge a t the donor site. This II back-donation leads to
a higher electron density on the two methyl hydrogens that are
trans to the lone-pair orbital. In fact, for methyl substitution athe
nitrogen donor site, Hehre and Pople40 have found that the two
methyl hydrogens out of plane with the lone pair become more
negatively charged. W e predict a similar, although attenuated
effect upon methyl substitution a t the a carbon atom. Due to
the strength of this interaction the three methyl hydrogens are
expected to be observably nonequivalent.
Interactions of the lone-pair orbital with the II-type filled and
unfilled orbitals of the methyl substituents at the a carbon atom
are hence predicted to change the electronic environment around
the nitrogen donor site relative to the unsubstituted case. An
energy diagram representing these interactions is shown in Figure
1. In summary, the interaction of the lone-pair orbital with the
IIMc*
unfilled methyl group orbitals (II back-donation) should
lead to a lower charge a t the donor site. As a result of this II
back-donation the methyl hydrogens trans to the lone-pair orbital
should have a higher electron density. The interaction of the
lone-pair orbital with the II-type filled methyl orbitals should result
in a slightly higher and more delocalized H O M O for the amino
species. However, this interaction will not lead to any significant
redistribution of electron density.
Translated into hard and soft acid-base theory, the effects of
methyl substitution a t th6 a carbon atom outlined above imply
that the amino species becomes a softer base upon such substitution. In terms of P M O theory, methyl substitution a t the a
carbon atom is predicted to enhance the relative importance of
the orbital-control term and reduce the importance of the
charge-control term for the interactions of the donor with a given
Lewis acid.
The arguments made so far predict that the interactions of the
nitrogen lone-pair orbital with the methyl group orbitals exhibiting
II-type symmetry alter the electronic environment a t the donor
site. The electronic properties of the molecule are altered such
that the donor has a higher and more delocalized H O M O with
a lower charge a t the donor site. The macroscopic implications
of the donor (RNH2) being a softer base in the reaction pathway
represented by steps 1-4 are now discussed.
Table 11. Some MNDO Results for Three Pairs of Amines
molecules compared
charge on nitrogen
HOMO energy
3.4% lower
1.6 kcal/mol higher
AMP compared to MEA
TBA compared to ETA
4% lower
1.2 kcal/mol higher
0.4 kcal/mol higher
TAM compared to PAM
4% lower
3. Implications for the Reaction Pathway
charge density
0.58
2
-0.01
1.01
3
-0.001
1.001
4
+0.015
0.98
-0.01
1.01
5
-0.001
1.001
6
Figure 2. Calculated partial charges on the various atoms of amines and
amino alcohols. Hydrogens on methyl substituents at the a-carbon atom
are not equivalent.
C 0 2 is a relatively hard acid; Le., the nu*
orbital that forms
its L U M O lies relatively high in energy. Therefore, the softer
the base RNH2, the weaker the N-C bond in the adduct formed
in the first step of the proposed reaction pathway. Consequently,
the depopulation of the IINHz
bonding orbitals should also be less.
If this depopulation is small enough, the step leading to the
formation of the carbamate species will be suppressed. On the
other hand, since OH- is hard base, the softer the base, RNH2,
the more step 2 is driven to the right. This enhances the formation
of the bicarbonate species. Thus, the predicted interactions of
the nitrogen lone-pair orbital with the methyl group orbitals that
lead to the base becoming softer upon methyl substitution a t the
a carbon atom, in turn result in suppressing the formation of the
carbamate species and enhancing the formation of the bicarbonate
species. This prediction is based only on the changes in the
electronic environment at the donor site upon methyl substitution
a t the a carbon atom; the possible changes in solvation energies
upon such substitution, however, are not incorporated. The
macroscopic kinetic and equilibrium observations of Chakraborty
et a1.’’J1 and Sartori and SavageI2 show that upon methyl substitution a t the a-carbon atom the formation of the carbamate
species is suppressed, and for all practical purposes the bicarbonate
species is the only product of the reaction (Le. a “chemical sink”
for COJ. The effects of methyl substitution on the intrinsic donor
properties of the amine and its consequent implications on the
various steps of the reaction pathway are therefore consistent with,
Table 111. Comparison of Energies of I I N ~Orbitals in
?
Methvl-Substituted and Unsubstituted
molecules compared
AMP compared to MEA
TBA compared to ETA
TAM compared to PAM
Amino Species
energy of IINHz
orbital
1 1.46 kcal/mol higher
9.96 kcal/mol higher
10.81 kcal/mol higher
-r.J
’?
TAM
atom no. charge charge density
-0.008
1
2
3
-0.007
+0.01
+0.01
-0.007
-0.008
4
5
6
TBA
atom no. charge charge density
1.009
1.007
0.989
0.989
1
2
3
4
5
6
1.007
1.009
-0.009
-0.007
+0.009
-0.008
-0.007
+0.009
1.009
1.007
0.99
1.008
1.007
0.99
1”
,A
atom no.
AMP
charge
1
+0.015
and serve to explain, the macroscopic kinetic and equilibrium
observations.
In order to substantiate the qualitative arguments made so far
and to place the discussion on a more quantitative basis, detailed
molecular orbital calculations have been performed. The results
of these molecular orbital calculations using the M N D O method7,42,43
and infrared spectroscopic experiments will now be
discussed.
4. Results and Discussion of MNDO Calculations
The M N D O computational procedure is a semiempirical method first proposed by Dewar and c o - ~ o r k e r s .The approxima~
(42) Dewar, M. J . S. The Molecular Orbital Theory of Organic Chemistry; McCraw-Hill: New York, 1969.
(43) Clarke, T. A Handbook of Computational Chemistry; Wiley: New
York, 1985.
J . Am. Chem. Soc., Vol. 110 No. 21, 1988 6951
.
Amine-C02 Interactions
Table IV. Comparison of Measured32 and Calculated (This Work)
amine
molecule
eth ylamine
n-propylamine
rerr-butylamine
rerr-amylamine
NH2
NP
ionizn potential,
kca I/ mol
measd
calcd
218.4
216.8
213.4
212.2
ORBITAL
TT
Ionization Potentials
deviation
kcal/mol
%
242.5
241.3
241.3
240.9
24.1
24.5
27.9
28.7
-11
-11
-13
-13
1s
Table V. MNDO Results Comparing HOCH2(OCH3)2NH2
and
CF,CH,NH, to Ethanolamine and Ethylamine, Respectively
molecules compared
HOMO energy
charge on nitrogen
CF3CH2NH2
compared to 20 kcal/mol lower
5% higher
ETA
3 kcal/mol lower
25% higher
HOH2C(OCH3)2CNH2
compared to MEA
tions embodied in the M N D O method are those inherent to the
N D D O f o r m a l i ~ m . ~ ~ * reported computational results are
The ~ ~
for completely optimized structures.
Tables I1 and 111 and Figure 2 show some of the relevant results
of the M N D O calculations performed for three pairs of amines
and aminoalcohols. The common structural feature relating these
three pairs is that tert-butylamine (TBA), tert-amylamine (TAM),
and 2-amino-2-methyl-] -propanol (AMP) are methyl substituted
at the a-carbon atom relative to ethylamine (ETA), n-propylamine
(PAM), and ethanolamine (MEA), respectively.
The results shown in Table I1 clearly indicate that upon methyl
substitution at the a-carbon atom the charge a t the nitrogen donor
site is lower. The H O M O of the amino species is less localized
and resides higher in energy for the methyl-substituted molecules.
It is worth noting that, in an attempt to correlate the proton
affinities of nitrogen and oxygen bases to their Is binding energies
and ionization potentials, Catalan et al." have reported the energies of the HOMOS of tert-butylamine and ethylamine. Catalan
and co-workers' STO-3G minimal basis set calculations also show
that the H O M O of tert-butylamine resides higher in energy than
that of ethylamine. Figure 2 depicts the calculated charges on
the atoms in 2-amino-2-methyl- I-propanol, tert-amylamine, and
tert-butylamine, These results show that the methyl hydrogens
that are trans to the lone-pair orbital have negative partial charges.
This provides evidence of II back-bonding; Le., the interaction
of the lone-pair orbital with the antibonding IIMe*
orbitals of the
methyl substituents. Thus, the II back-donation observed by
Hehre and Poplea upon methyl substitution at the nitrogen donor
site is also important upon methyl substitution one carbon atom
removed. The partial charges on the a-carbon atom for methyl-substituted molecules are calculated to be negative as compared
to the positive charges in the unsubstituted molecules. This is
perhaps due to the fact that in addition to the lone-pair orbital,
the a-carbon orbitals also interact with the filled I orbitals of
I
the methyl substituents. The possibility of such interactions in
methyl-substituted carbon acids has been considered by Pross and
Rad~m.~~
Aue and co-workers3*have reported the gas-phase ionization
potentials for four of the amino species investigated in this work
(ETA, TBA, PAM, and TAM). In Table IV the ionization
potentials measured by Aue and
using photoelectron
spectroscopy are compared with the ionization potentials calculated
in this work using the M N D O method and Koopmans' theorem.42
Table IV shows that the trend in the variation of the ionization
potential with structure is reproduced well. The absolute values
of the ionization potentials, however, are not reproduced exactly
by the calculations (average deviation 12%). Furthermore, the
results shown in Table IV show that the lowering of the ionization
potential (or raising of the H O M O ) upon methyl substitution a t
the a-carbon atom is underestimated by the M N D O calculations.
-
(44) Catalan, J.; Mo, 0.; Perez,
Trans. 2 1982, 1415.
IS
Figure 3. Simplest possible representation of the IINHI.
bonding orbital.
For methyl-substituted systems this orbital is destabilized by about 10
kcal/mol.
This is perhaps to be expected since the M N D O method is known
to underestimate hyperconjugative effects.43 The M N D O predictions of the effects of orbital interactions considered in this paper
are thus lower limits. Aue and co-workers have performed
C N D O / 2 calculations on unoptimized structures in order to
calculate the proton affinities of amino species. Their calculations
do not reproduce the experimental proton affinities well (50% too
high). Aue et al.32have rationalized their experimental data via
polarizability stabilization arguments.
The hardness parameter, 7, introduced by Parr and P e a r ~ o n ~ ' - ~ ~
can be calculated from the M N D O results by using Koopmans'
theorem; Le. by approximating the ionization potential as -EHOMO
and the electron affinity as -ELUMO. The parameter, 7, reflects
the HOMO-LUMO energy separation. The M N D O calculations
carried out in this work show that, on an average, the HOMOLUMO energy gap is reduced by about 4.5 kcal/mol upon methyl
substitution a t the a-carbon atom. This lowering of the
HOMO-LUMO energy gap provides further evidence that the
base becomes softer upon methyl substitution a t the a-carbon
atom.
A measure of the strength of the N-H bonds in a primary
amino species is the energy of the I I N H 2 bonding orbitals. In its
simplest representation the n N H 2 bonding orbital is obtained by
a constructive combination of the hydrogen Is orbitals with a
nitrogen p orbital. This simple representation of the I I N H 2 bonding
orbital is depicted in Figure 3. Of course, in any given molecular
environment there are contributions from other atomic orbitals
to this simple representation. The results of the M N D O calculations reported in Table 111 show that the I I N H 2 bonding orbital
is destabilized by about 10 kcal/mol upon methyl substitution a t
the a-carbon atom. This relative destabilization may be explained
as follows. For the unsubstituted molecules, there is a u contribution to the I I N H 2 bonding orbital from the hydrogens on the
a-carbon atom. However, for the methyl-substituted molecules
there is a u contribution from the p orbitals of the carbons of the
methyl substituents. The overlap with the p orbitals in this u
interaction is smaller, and hence the I I N H 2 orbital is destabilized
upon methyl substitution.
On the basis of ion cyclotron resonance experiments Brauman
and Blair45have reported that amines with large alkyl groups are
more acidic. They4s propose the steps in eq 8-10 for the loss of
A H = A'
+ H'
A ' + e- = A-
H' = Hf
+ e-
DH,
(8)
-A
(9)
I of H
(10)
a proton where DHo is the dissociation energy of the species, AH.
Brauman and Blair45 have ascribed the relative ease of proton
abstraction from amines that have large alkyl substituents to
polarizability stabilization of the charge on the anion by large alkyl
groups. It is to be noted that charge stabilization changes the
electron affinity (corresponding to step 9). The results reported
in Table 111 show that the dissociation energy (corresponding to
step 8) is also lowered upon methyl substitution a t the a-carbon
atom. This change in the dissociation energy has been observed
P.; Yanez, M. J . Chem. Soc., Perkin
(45)
Brauman, J. I.; Blair, L. K. J .
Am.
Chem. SOC.1971, 93, 3911.
Chakraborty et al.
6952 J. Am. Chem. SOC.,
Vol. 110, No. 21, 1988
w
R
3.5 Ao
Figure 4. Calculations along the reaction coordinate for the C02-tert-
butylamine interaction. The C-N distance (reaction coordinate) is 3.5
A.
1.75 A'
Figure 5. Calculations along the reaction coordinate for the COz-tertbutylamine interaction. The C-N distance (reaction coordinate) is 1.75
A.
experimentally4' for the series NH3, CH3NH2,and (CH3)2NH.
The results reported here show that methyl substituents a t the
a-carbon atom also weaken the N-H bond within a series of
primary amino compounds.
In an effort to explore the reaction coordinate for COz-amine
interactions, M N D O calculations were performed a t various
amine-COZ separations for ethylamine and tert-butylamine. The
distance between the carbon atom in C02and the nitrogen donor
site of the amino species was chosen as the reaction coordinate.
Figures 4 and 5 show the results of these calculations driving along
the reaction coordinate for the tert-butylamine-CO* interaction.
Figure 4 shows the optimized configuration when the C-N reaction
coordinate distance is 3.5 A. At this point the COz molecule is
still linear, and no significant interaction appears to have occurred.
Figure 5 shows the optimized configuration at another point along
the reaction coordinate. This figure clearly shows how the C 0 2
molecule folds toward a bent structure. The important features
to notice in Figure 5 are the folding of the C02 molecule as it
approaches the amino species and that there seem to be no obvious
steric constraints imposed by the methyl groups on the a-carbon
atom. In an effort to quantify possible steric effects, the energy
barrier to the rotation of the COz molecule about the C-N axis
in an approximate transition state was calculated for both ethylamine and tert-butylamine. This rotation was found to be almost
equally facile for both amines. This allows one to conclude that,
with respect to this rotational mode, there appears to be little
change in steric constraints due to the methyl substituents at the
a carbon atom.
The observed instability of the carbamate species has been
explained by the fact that since the base becomes softer upon
methyl substitution a t the a-carbon atom, the N-C bond in the
adduct is not strong enough to lead to the loss of a proton. M N D O
calculations show this change in the electronic environment at the
donor site upon methyl substitution a t the a-carbon atom. The
relative instability of the carbamate species could be further
investigated computationally by calculating the overall stabilization
of step 1 in the reaction pathway. To calculate this overall heat
of reaction, step 1 may be represented as eq 11. The calculated
RNH2
+ COZ = R N H C O O H
differences in the heats of reaction (-AH) for the overall reaction
Table VI. Overall Stabilization Energies and C-N Orbital Energies
in Ethylcarbamic Acid and tert-Butylcarbamic acid
-AH,
amine interacting
with CO,
kcal/mol
ETA
TBA
8.6
5.1
C-N orbital energy, eV
-20.36
-19.46 (20.7 kcal/mol higher)
(eq 11) for different substituents, R, serve as a rough indicator
of the relative stabilities of the carbamate species. Furthermore,
the energy of the C-N bonding orbital in the carbamic acid would
be indicative of the strength of the interaction. Calculations
performed for ethylamine and tert-butylamine show that eq 11
leading to the formation of ethylcarbamic acid is more favorable
than that leading to the formation of tert-butylcarbamic acid
(Table VI). This provides further evidence for the intrinsic
instability of the carbamate species upon methyl substitution a t
the a carbon atom. The predominantly u-type C-N bonding
orbital is also found to reside about 20 kcal/mol higher in energy
for tert-butylcarbamic acid compared to ethylcarbamic acid, thus
offering additional evidence that the higher and more delocalized
H O M O and the lower charge at the donor site leads to weaker
amine-C02 interactions.
In an effort to further quantify any steric effects upon methyl
substitution a t the a carbon atom, the energy barriers for the
rotation of the carboxylic group about the C-N bond in ethylcarbamic acid and tert-butylcarbamic acid have been calculated.
These calculations were performed by allowing the rest of the
molecule to fully relax in the various fixed rotational configurations. At the M N D O level of theory, the difference in the energy
barrier for rotation of the -COOH group between ethylcarbamic
acid and tert-butylcarbamic acid is found to be 0.7 kcal/mol. This
small difference suggests again that, with respect to this rotational
mode, steric effects are not significantly altered upon methyl
substitution a t the a-carbon atom. Further detailed studies need
to be conducted to preclude the presence or absence of steric effects
with respect to other modes. Furthermore, methyl substituents
a t the donor site, or bulkier substituents a t the a-carbon atom
may lead to steric limitations.
It is interesting to investigate the effects of methoxy or trifluoromethyl substitution a t the a-carbon atom on the chemistry
of primary amino donors. In these cases qualitative arguments
suggest that inductive effects should result in the base becoming
harder with a higher charge a t the donor site. However, the
possibility of interactions with the methyl moieties of the methoxy
substituents cannot be ruled out a priori. Table V shows the results
of M N D O calculations performed for the doubly methoxy-substituted amino alcohol and 2,2,2-trifluoroethylamine. (The
methoxy-substituted amino alcohol has not been reported in the
literature. However, the purpose of this calculation is to examine
the effects of polar substituents of this type on the donor properties
and its concomitant implications for acid gas-separation processes.)
These results show that the base does become harder upon such
substitutions. This observation suggests that perhaps this is the
kind of amine or amino alcohol that should be used to separate
an acidic gas such as SO2 that has a soft acid site. For the
methoxy-substituted molecule, no evidence of II back-bonding can
be seen at the M N D O level of theory. Thus, the results reported
in this work, along with those of Hehre and Pople,40 show that
while the interaction of the lone-pair orbital with the nMe*
antibonding methyl group orbitals is important for methyl substitution a t the heteroatom as well as the a-carbon atom, it is unimportant when the methyl moieties are two atoms removed. This
is due to the negligible overlap in this case.
The M N D O calculations show that, upon methyl substitution
a t the a-carbon atom, the nitrogen lone-pair orbital interacts with
the II-type filled and unfilled methyl group orbitals. This changes
the donor properties of the amino species. In particular, it leads
to a lower charge at the donor site and a less localized and higher
energy HOMO. The calculated results also show the relative
antibonding
importance of lone-pair interactions with the nMC*
methyl orbitals with changes in the substitution site. The IINH2
bonding orbitals of the amine are also shown to be destabilized
J . Am. Chem. SOC..Vol. 110, No. 21, 1988 6953
Amine-C02 Interactions
Table VII. Frequencies of the N-H Asymmetric Stretching
Vibration for n-Propylamine and tert-Amylamine in CC14 solution
amine
N-H asymmetric stretching mode, cm-I
PAM
3391
TAM
3378
Table VIII. Band Assignments of C-H Stretching Vibrations of
n-Propylamine and tert-Amylamine in CC14 Solution
amine
Y, cm-I
assgnt
PAM
(1) 2960
asym -CH3
(2) 2875
sym -CH,
(3) 2932
asym -CH2
(4) 2856
sym -CH2
TAM
(2) 2937
-CH3
(3) 2879
( I ) 2963
(4) 2923
asym -CH2
(5) 2854
sym -CH2
I
r
,
4910
4630
4350
4070
3790
I
3510
3230
2950
2670
2390
WAVENUMBERS
Figure 6. Resolved infrared spectra showing the bonded -OH bands for
the interactions of terf-amylamine (TAM) and n-propylamine (PAM)
with phenol.
upon methyl substitution a t the a-carbon atom.
It is appropriate a t this point to discuss the results of some
infrared spectroscopic experiments that further substantiate the
MNDO results regarding the changes in the donor properties of
aliphatic amino species upon methyl substitution a t the a carbon
atom.
5. Infrared Experiments
The infrared experiments reported in this work were performed with
spectral-grade reagents without further purification. All samples were
prepared in a dry glovebox with thoroughly dried glassware. The spectra
were obtained on a Nicolet 7000 series Fourier transform infrared
spectrometer at a resolution of 0.24 cm-l.
The infrared spectra of the amines in carbon tetrachloride were obtained at room temperature (21 "C). Experiments were performed over
a wide range of concentration levels in order to ascertain that association
effects were eliminated. Table VI1 shows the asymmetric stretching
frequencies of the N-H bonds of tert-amylamine (TAM) and npropylamine (PAM) in CCI4 at very high dilution. These values are
consistent with Kreuger and Smith'su investigations of the N-H
stretching frequencies of primary amino species. Structurally, the two
amines are different in that rert-amylamine is methyl substituted at the
a-carbon atom compared to n-propylamine. The frequencies reported
in Table VI1 indicate that the N-H bond is weakened upon methyl
substitution at the a-carbon atom. This result is consistent with MNDO
calculations reported earlier that show that the IINH2
orbital is destabilized upon methyl substitution at the a-carbon atom. As discussed
earlier, this change in the N-H bond strength can be related to the
changes in the acidity of the amino species observed in somewhat similar
systems by Brauman and Blair.45
Table VI11 reports the frequencies of the observed C-H stretching
region for tert-amylamine and n-propylmodes in the 2850-3000-~m-~
amine. Table VI11 indicates that the band pattern in this region is
different for the two amines. In particular, the spectrum of tert-amylamine shows the existence of a band at 2937 cm-' in addition to the four
bands corresponding to the standard asymmetric and symmetric
stretching modes of methyl and methylene groups. Several workers in
the past47-50
have observed the appearance of an extra band in the C-H
stretching region when methyl groups are in the vicinity of a lone pair.
The existence of this extra band was initially ascribed to Fermi resonance
with one of the C-H deformation modes."* However, later workSohas
shown that an interaction with the lone pair is the probable cause. The
MNDO calculations reported earlier in this paper show that upon methyl
substitution at the a-carbon atom, the 11-type lone-pair orbital interacts
with the
antibonding methyl orbitals. This leads to a higher
electron density on the two methyl hydrogens that are trans to the lone
pair. Consequently, the three hydrogens are not equivalent, and more
than two bands for the C-H stretching modes of the methyl groups in
tert-amylamine should be observed. On the basis of this argument the
(46) Kreuger, P. J.; Smith, D. W. Can. J . Chem. 1967, 45, 1605.
(47) N o h , B.; Norman Jones, R. J . Am. Chem. SOC.1953, 75, 5626.
(48) Bellamy, L. J. Advances in Infrared Group Frequencies; Methuen:
London, 1968.
(49) Higuchi, S.; Kuno, E.; Tanaka, S.; Kamason, H. Spectrochim. Acta,
Part A 1972, 28A, 1335.
(50) Bellamy, L. J. The Infrared Spectra of Complex Molecules; Chapman and Hall: London, 1975.
PAM
ire,
z
u
NH
OH
4
m
rr
TAM
, \
3700
3550
3400 3250
3100
2950
2800 2650
2500
I
2350
WAVE N UMBER S
Figure 7. Resolved infrared spectra showing the bonded -OH bands for
the interactions of tert-amylamine (TAM) and n-propylamine (PAM)
with p-chlorophenol.
extra band may be tentatively assigned to the methyl groups attached to
the a-carbon atom. Currently, partial deuteriation studies are under way
in this laboratory to ascertain this assignment.
The infrared spectra of tert-amylamine and n-propylamine thus indicate that upon methyl substitution at the a-carbon atom the N-H bond
is weakened and that the interaction of the lone-pair orbital with the
IIMF*
methyl orbitals leads to the three methyl hydrogens being inequivalent.
A major research effort by the Drago g r o ~ p ~has ~ ~
' - established that
the frequency shifts of the -OH band (AYoH) of phenolic acids upon
complexation with nitrogen and oxygen donors can be linearly related to
the bond strength in the adduct. With nitrogen or oxygen donors phenolic acids form a rather strong hydrogen bond. This leads to a weakening of the adjacent -OH bond in the phenolic acid. This consequently
results in a broadening and downfield shift of the band corresponding to
the -OH stretching mode. The Drago group's work establishes that this
downfield shift is directly related to the strength of the hydrogen bond.
Hence, by carrying out complexation reactions of a series of well-characterized phenolic acids with a series of bases, one could ascertain the
relative hardness or softness of the donor species. In particular, the
changes in the donor properties of amino species upon methyl substitution
at the a-carbon atom could be characterized by performing such complexation studies using infrared spectroscopy. The results of complexing
a pair of amines, tert-amylamine, and n-propylamine with phenol and
p-chlorophenol are reported here. It has been established by Drago and
co-workerss3 that p-chlorophenol has the harder acid site compared to
phenol. Thus, on the basis of the results of complexing n-propylamine
and its methyl-substituted analogue, tert-amylamine with these two
phenolic acids, one should be able to conclude about the relative hardness
or softness of the two donors.
(51) Drago, R. S.; Epley, T. D. J . Am. Chem. SOC.
1969, 91, 2883.
1971,
(52) Drago, R. S.; Vogel, G . C.; Needham, T. E. J . Am. Chem. SOC.
93, 6014.
(53) Purcell, K. F.; Drago, R. S. J . Am. Chem. SOC.
1967, 89, 2874.
(54) Joesten, M. D.; Drago, R. S. J . Am. Chem. SOC.
1962, 84, 2037.
Chakraborty et al.
Vol. 110, No. 21, 1988
6954 J . Am. Chem. SOC.,
The phenolic acid-amine complexes were prepared by adding amine
directly to the phenolic acid and then diluting with the solvent, carbon
tetrachloride. Enough amine was added to complex about 90% of the
phenolic acid. As in the case of the amine spectra, experiments were done
over a wide range of concentration levels to eliminate association effects.
The reactions were carried out at room temperature (21 “C), and the
spectra were also recorded at room temperature.
Figure 6 depicts the resolved spectra obtained on complexing the two
amines (represented as TAM and PAM) with phenol. The free-OH
band is shown for reference. All other bands corresponding to the amine
and phenol have been subtracted out. The bands labeled N-H are the
downfield shifted N-H asymmetric stretching bands corresponding to the
complexed amines. The broad band is the downfield shifted-OH band
for the complexed phenol. Figure 7 shows similar spectra for the interactions of the two amines with p-chlorophenol.
The differences in AuoH and peak width of the complexed -OH bands
for the four interactions depicted in Figures 6 and 7 are indeed interesting. Figure 6 shows that for the interactions with phenol, the bonded
-OH band is more downfield shifted and has a greater peak width in the
phenol-terr-amylamine complex. This implies that with the softer acid,
phenol, the methyl-substituted rerr-amylamine forms the stronger bond.
In contrast, Figure 7 shows that with p-chlorophenol the bonded -OH
band is more downfield shifted and has a greater bandwidth in the pchlorophenol-n-propylamine complex. This implies that, with p-chlorophenol, which has the harder acid site, n-propylamine forms the stronger
bond. Hence, these data allow us to conclude that the donor amine does
become softer upon methyl substitution at the a-carbon atom. The
complexation studies are thus consistent with the MO arguments presented in this work.
Drago and c o - w o r k e r ~and KlopmaniPhave suggested the use of
~~*~~
the E & C equation along with the AvoH values to further quantify the
relative hardness or softness. However, since the peak width of the
bonded -OH bands are large, it has not been deemed fit to quantify the
relative softness of the systems studied in this work in this manner.
The differences in donor properties upon methyl substitution, as evidenced by the molecular orbital calculations and the infrared studies,
rationalize some of the macroscopic observations of amine-CO, reactions
in solution.*’* Of course, it is recognized that differences in solvation
effects could be important. However, changes in the intrinsic chemistry
alone (without accounting for solvation effects) predict the trends in the
macroscopic behavior upon methyl substitution at the a-carbon atom.
While it is true that for the ionic species such as the protonated amine
and the carbamate anion differences in solvation energies upon methyl
substitution could be important, in the cases under investigation the
interaction of the free amine with C 0 2 is of primary consideration.
Hence, one needs to focus on the possible differences in solvation energies
of the free amine itself upon methyl substitution at the a-carbon atom.
It is therefore appropriate to briefly examine some of the possibly reasons
for differences in solvation energies of the free amines upon such substitutions.
The amino species considered in this work are a series of primary
amines or amino alcohols. Hence, the number of hydrogen bonds that
can be formed is the same for substituted and unsubstituted molecules.
Thus, differences in amine solvation energies due to different degrees of
hydrogen bonding should not be significant. The difference in the partial
charge at the nitrogen donor site upon methyl substitution at the a(55) Drago, R.
s.;Wayland, B. B. J . Am. Chem. SOC.1965, 87, 3571.
carbon atom could contribute to different solvation energies. Calculations
using the Born e q ~ a t i o n , ~however, predict that for the charge dif~.~’
ferences reported in this paper this effect should be small. Steric inhibition to solvation due to relatively small methyl groups at the a-carbon
atom should also not be significant.” These qualitative arguments suggest that the change in the solvation energy of the free amine upon
methyl substitution at the a-carbon atom should not be significant. This
is consistent with the Trotman-Dickenson proposal.58 Arnett and cow o r k e r have reported that the difference in the free energy of solvation
~~~
between ethylamine and ferf-butylamine is only 0.24 kcal/mol. This
experimental result supports the qualitative arguments suggesting that
the difference in solvation energies is not significant for the free amine
and hence its interactions with C02. Furthermore, as noted recently by
P e a r s ~ nif, ~ ~ net charge on a molecule or pair of interacting molecules
the
does not change significantly, then the hardness parameter as determined
from the gas-phase ionization potential and electron affinity is a fairly
good approximation even in solution. The extension of the molecular
orbital arguments, the MNDO results, and the infrared studies toward
rationalizing the macroscopic behavior in a qualitative sense is thus
justified. Development of quantitative structure-activity relationships
will, however, require a detailed accounting of solvation effects.
6 Concluding Remarks
.
Qualitative molecular orbital arguments, M N D O calculations,
and infrared data reported in this paper demonstrate that methyl
substitution at the a-carbon atom in primary amines and amino
alcohols leads to subtle but significant changes in the electronic
environment of the donor site, nitrogen. These changes are due
to the interactions of the nitrogen lone-pair orbital with the IIMc
and nMe*
methyl group orbitals. These interactions result in the
donor species having a lower charge at the donor site and a higher
and more delocalized H O M O (Le., making it a softer base).
Furthermore, the N-H bond of the amine is weakened upon
methyl substitution a t the a-carbon atom.
The results reported in this paper show that the differences in
the donor properties of amino species upon methyl substitution
at the a-carbon atom should lead to the experimentally observed
differences in reactivity and selectivity in amine-C02 reactions.
Since the effects of the orbital interactions considered in this paper
are probably underestimated by M N D O calc~lations,4~ initio
ab
calculations have recently been carried out in this research group
to further study the orbital interactions discussed in this work.
These results will be reported in an ensuing publications9 and show
that a b initio calculations using extended basis sets predict the
effects reported in this paper to be further accentuated.
Registry No. Amp, 124-68-5;
MEA, 141-43-5;
TBA, 75-64-9;
ETA,
75-04-7;
TAM, 594-39-8;
PAM, 107-10-8;
CO,, 124-38-9;
CF$HZNH,,
753-90-2;
HOH2C(OCH3),CNH2, 11551 5-23-6;
ethylcarbamic acid,
7409-13-4;
terf-butylcarbamic acid, 69777-55-5;
phenol, 108-95-2;
pchlorophenol, 106-48-9.
(56)Born, M. Z.
Phys. 1920, 1 , 45.
(57)Klopman, G.Chem. Phys. Left. 1967, 1 , 200.
(58) Trotman-Dickenson, A. F. J . Chem. SOC.
1949, 1293.
(59)Chakraborty, A. K.;
Bischoff, K. B.; Astarita, G.; Damewood,J. R.,
to be submitted for publication in Chem. Phys. Letf.