pH dependence of the electronic structure of glycine.

Abstract

The carbon, nitrogen, and oxygen K-edge spectra were measured for aqueous solutions of glycine by total electron yield near-edge X-ray absorption fine structure (TEY NEXAFS) spectroscopy. The bulk solution pH was systematically varied while maintaining a constant amino acid concentration. Spectra were assigned through comparisons with both previous studies and ab initio computed spectra of isolated glycine molecules and hydrated glycine clusters. Nitrogen K-edge solution spectra recorded at low and moderate pH are nearly identical to those of solid glycine, whereas basic solution spectra strongly resemble those of the gas phase. The carbon 1s --> pi*(C=O) transition exhibits a 0.2 eV red shift at high pH due to the deprotonation of the amine terminus. This deprotonation also effects a 1.4 eV red shift in the nitrogen K-edge at high pH. Two sharp preedge features at 401.3 and 402.5 eV are also observed at high pH. These resonances, previously observed in the vapor-phase ISEELS spectrum of glycine, have been reassigned as transitions to sigma* bound states. The observation of these peaks indicates that the amine moiety is in an acceptor-only hydrogen bond configuration at high pH. At low pH, the oxygen 1s --> pi*(C=O) transition exhibits a 0.25-eV red shift due to the protonation of the carboxylic acid terminus. These spectral differences indicate that the variations in electronic structure observed in the NEXAFS spectra are determined by the internal charge state and hydration environment of the molecule in solution.

Full text

Page 1

pH Dependence of the Electronic Structure of Glycine
B. M. Messer,†C. D. Cappa,†J. D. Smith,†K. R. Wilson,‡M. K. Gilles,‡R. C. Cohen,†and
R. J. Saykally*,†
Department of Chemistry, UniVersity of California, Berkeley, California 94720-1460, and
AdVanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, California 94720
ReceiVed: September 17, 2004; In Final Form: NoVember 15, 2004
The carbon, nitrogen, and oxygen K-edge spectra were measured for aqueous solutions of glycine by total
electron yield near-edge X-ray absorption fine structure (TEY NEXAFS) spectroscopy. The bulk solution pH
was systematically varied while maintaining a constant amino acid concentration. Spectra were assigned through
comparisons with both previous studies and ab initio computed spectra of isolated glycine molecules and
hydrated glycine clusters. Nitrogen K-edge solution spectra recorded at low and moderate pH are nearly
identical to those of solid glycine, whereas basic solution spectra strongly resemble those of the gas phase.
The carbon 1s f π*CdOtransition exhibits a 0.2 eV red shift at high pH due to the deprotonation of the
amine terminus. This deprotonation also effects a 1.4 eV red shift in the nitrogen K-edge at high pH. Two
sharp preedge features at 401.3 and 402.5 eV are also observed at high pH. These resonances, previously
observed in the vapor-phase ISEELS spectrum of glycine,1have been reassigned as transitions to σ* bound
states. The observation of these peaks indicates that the amine moiety is in an acceptor-only hydrogen bond
configuration at high pH. At low pH, the oxygen 1s f π*CdOtransition exhibits a 0.25-eV red shift due to
the protonation of the carboxylic acid terminus. These spectral differences indicate that the variations in
electronic structure observed in the NEXAFS spectra are determined by the internal charge state and hydration
environment of the molecule in solution.
1. Introduction
Individual amino acids exist as cations in acidic media, charge
neutral zwitterions at intermediate pH values, and anions in basic
solutions. This strong pH dependence of the charge state is
reflected in the sensitivity of properties of the peptides and
proteins built from individual amino acids to changes in their
environment. Such changes in local environment can alter
protein shape and reactivity, which in turn can have dramatic
effects on their biological activity.2Hence, the electronic
structure and chemical behavior of amino acids has been of
continued interest to both chemists and molecular biologists.
Soft X-ray spectroscopy has been used extensively to character-
ize electronic structure in amino acids and peptides.1,3-11Near-
edge X-ray absorption fine structure (NEXAFS) spectroscopy,
inner shell electron energy loss spectroscopy (ISEELS), and
scanning transmission X-ray microscopy (STXM) have been
employed in studies of spin- cast thin films of all 20 basic amino
acids,6amino acid monolayers adhered to metallic substrates,3,4
and both vapor9-11and solid thin film samples of specific amino
acids and their respective di- and tripeptides.1,7,8These experi-
ments have been supported by a variety of theoretical ef-
forts.3,4,6,12,13
Because of technical limitations, all of these studies have been
performed on amino acids in nonbiological environments.
Biological systems are almost universally aqueous systems.2
Isolated amino acids are strictly neutral in the gas phase but
take the form of zwitterions in the condensed phase. However,
in solution, a wide variety of charge states exist, the relative
populations of which are determined by the pH of the solution.
If these differences in charge state were caused simply by the
addition or removal of an electron, one might expect that the
NEXAFS spectra of amino acids would be relatively insensitive
to changes in pH. However, the addition or removal of a proton
alters the local symmetry of the electric field surrounding the
terminal nitrogen or oxygen. Thus, a change in pH can engender
a dramatic difference in the energy levels of any molecular
orbitals involving the amine or carboxylic acid moieties. Recent
studies by Gordon et al.1and Zubavichus et al.5invoke this
interpretation to explain the disappearance of resonances in the
oxygen K-edge spectrum of glycine observed upon condensa-
tion.
The nitrogen K-edge spectrum also exhibited dramatic
differences between solid and gas-phase glycine.1Two of the
sharp resonances in the vapor phase spectrum have been
assigned to Rydberg states, and Gordon et al. rationalize the
disappearance of these peaks in the thin film spectrum by noting
that gas-phase NEXAFS spectra have sharp resonances that
often either broaden into a continuum or disappear altogether
upon condensation. However, given the change in charge state
that occurs upon condensation, their interpretation is not unique.
In aqueous solutions of glycine, both solvent and solute
molecules surround each amino acid molecule. Therefore, if
condensation effects were responsible for the observed spectral
variations, the nitrogen K-edge spectrum of aqueous glycine
should exhibit little pH dependence. A strong pH dependence
in the nitrogen K-edge would instead suggest that charge state
and solvation environment dominate the behavior of the
NEXAFS spectrum.
Kaznacheyev et al.6have previously attempted to measure
the pH dependence of the carbon K-edge using STXM, although
* Corresponding author. E-mail: saykally@cchem.berkeley.edu.
†University of California.
‡Lawrence Berkeley National Laboratory.
5375
J. Phys. Chem. B 2005, 109, 5375-5382
10.1021/jp0457592 CCC: $30.25 © 2005 American Chemical Society
Published on Web 01/22/2005

Page 2

the results of those measurements were ambiguous, as the small
liquid droplets sandwiched between silicon nitride windows used
in that study yielded insufficient sensitivity. Ultimately, the
carbon K-edge is unlikely to exhibit any strong pH dependence,
as the solid and vapor spectra are very similar to each other,
and little change in the electronic structure is expected as the
charge state is varied. Additionally, there is some uncertainty
as to the charge state of the solid thin films used in previous
STXM studies.6To produce the amorphous thin films appropri-
ate for STXM measurement, the amino acid samples were first
dissolved in trifluoroacetic acid (TFA). The TFA then rapidly
evaporates, and the acidity of the resulting sample is assumed
to be the same as that of the original material.
At least two other techniques have also been employed to
determine the pH dependence of amino acid and protein
solutions. NMR spectroscopy has been used to investigate the
folding and interchain coupling of numerous proteins as a
function of pH.14-16Additionally, NMR studies have been used
to determine the number of water molecules participating in
hydrogen bonding with either the amine or carboxylic acid
portions of amino acids.17,18Raman spectroscopy studies have
been used to determine when polypeptide chains first start to
form R-helixes,19to determine the helical content of various
proteins as a function of pH,20and along with vibrational circular
dichroism (VCD) were used to measure the optical activity of
alanine21and asparagine22as a function of pH. NEXAFS differs
from NMR and Raman spectroscopies in that it is a direct probe
of electronic structure.23Ideally, our understanding of the
electronic structure of proteins and complex biomolecules could
be enhanced from a thorough description of the electronic
structure of the constituent amino acids and polypeptides.
It is the goal of this work to improve such understanding by
systematically exploring the effect of solvation and pH on amino
acid electronic structure. Here we report the experimental
measurement of the NEXAFS spectra of aqueous glycine for
all relevant charge states and elemental K-edges. To overcome
the technical difficulties inherent in characterizing volatile
liquids, we have exploited the technology of liquid microjets
originally developed by Faubel et al.24and further developed
by Wilson et al.25-27Extensive density functional theory studies
were performed using a commercial software package28to make
spectral assignments and analyze the effect of pH on the
electronic structure of aqueous glycine. To our knowledge, this
is the first such study of this system performed in an aqueous
environment.
2. Experimental and Computational Methods
2.1. Samples. Glycine (C2H5NO2, NH2CH2CO2H) was ob-
tained commercially from Fischer Scientific in the form of a
crystalline powder with a stated purity of better than 99%. This
was used without further purification. Sample solutions were
prepared by dissolving enough glycine in 18 MΩ water
(Millipore) to create a 10% by mass glycine solution. To create
acidic or basic samples, an equimolar amount of reagent grade
HCl or NaOH, also obtained from Fischer, were added to the
crystalline glycine before being diluted to 10% glycine with 18
MΩ water. These solutions were later mixed with equal volumes
of deionized water to form an approximately 5% by mass (∼0.6
M) solution.
2.2. NEXAFS Spectroscopy of Aqueous Solutions. The
X-ray absorption spectra were recorded over all pertinent energy
ranges for each sample solution. The carbon K-edge (∼300 eV),
nitrogen K-edge (∼400 eV), and oxygen K-edge (∼530 eV)
were recorded at Beamline 8.0.1 of the Advanced Light Source
in Berkeley, CA. A detailed description of the experimental
apparatus has recently been given by Wilson et al.27Briefly,
an intense (1012photons s-1) tunable beam of X-ray radiation
is generated by means of a undulator insertion device. This
spectrally narrow (∆E/E ) 4000 at 500 eV) radiation is focused
onto an approximately 30 µm diameter jet of the sample
solution. The jet is produced by an HPLC pump (Jasco PU-
2089) pressurizing the liquid to ∼5 MPa behind a fused glass
capillary tip pulled with a CO2laser to a final diameter of 30
µm. This jet is then injected horizontally into an evacuated
chamber, parallel to the polarization vector of the incident
radiation. After traveling briefly (∼2 mm ) ∼80 µs) in a
vacuum, the jet is intersected by a 50 µm diameter X-ray beam
at 90°. The room temperature solution evaporatively cools upon
injection into the vacuum chamber, and the sample is determined
to be 15-20 °C27when probed by the X-ray beam. The jet is
subsequently collected in a separate cryogenic trapping chamber.
The liquid microjet approach provides a continuous supply of
fresh sample solution, avoiding the introduction of contamination
from radiation damage. The probability of a molecule being
doubly excited during the ∼5 µs exposure time is calculated to
be approximately 10-7.
The pressure in the interaction chamber is maintained at ∼6
× 10-5Torr, whereas two sections of differential pumping allow
windowless coupling to the beamline, maintained at ∼5 × 10-10
Torr. This windowless coupling permits the measurement of
all three edges without any interference or loss of flux introduced
by a Si3N4window, used in the previous NEXAFS studies of
amino acids and poly-peptides.1,5-8The decreased background
pressure is sufficiently low to allow efficient collection of the
electrons created by the X-ray excitation of the sample. In this
way, the total electron yield (TEY) is sampled by means of a
positively biased 1 mm2copper electrode placed within 0.5 mm
of the interaction region. The detected current is amplified and
converted to a voltage before being sent to a V-F converter
for input into the beamline computer. This signal is then
normalized to the current measured from a gold mesh located
3 m upbeam of the interaction region. Specific metal impurities
in the monochromator optics provide an X-ray energy calibration
over the entire range of interest. For gas-phase systems, observed
line widths in the energy range 200-500 eV are generally
determined by the lifetime of the respective core hole decay
processes and the lifetime of the resulting photoelectron. These
natural line widths are inhomogeneously broadened in the
condensed phase because of the increased density of states and
solvent interactions.
2.3. Computational Methods. To analyze the spectra col-
lected in these experiments, the NEXAFS spectra of each charge
state of glycine were calculated using density functional and
transition-potential methods. The StoBe DeMon 2.028software
package is a commercially available implementation of these
approaches, and was used for all energetic and spectral
calculations presented in this work. In these calculations, the
correlation functional formulated by Perdew, Berke, and Ern-
zerhof29and the revised exchange functional formulated by
Hammer30were used along with the optimized DZVP basis set
provided. The single point energies and optimized geometries
were calculated for several conformers of each glycine charge
state under investigation. The resulting energy of each conformer
was then used to generate a Boltzman distribution for each
charge state. For each species, only one conformer was found
to be significantly populated at the jet temperature (15 °C).
These geometries, depicted in Figure 1, were then used for all
5376 J. Phys. Chem. B, Vol. 109, No. 11, 2005
Messer et al.

Page 3

further calculations. For reference, the Cartesian coordinates of
each atom are provided in Table 1.
To calculate the NEXAFS spectra of each conformer, the
minimum-energy geometries and exchange/correlation func-
tionals were maintained while the basis sets were changed as
follows. The atom of interest was modeled using the IGLO basis
set developed by Huzinaga,31and all remaining non-hydrogen
atoms were replaced with the prepackaged model core potentials
developed by Pettersson.28This allows the energetic isolation
of the 1s state of interest, which is then partially excited by
replacing its R electron with a half core hole. This “transition-
potential” method has been used in multiple R scattering
calculations and DFT calculations to simulate the excited state
of the system.23The transition-potential method has been found
to approximate the relaxation of the molecule under the influence
of the core-hole and is sufficiently accurate and self-consistent
for the relative comparisons made in this work. The resulting
list of transitions and oscillator strengths are then convoluted
with a series of Gaussians to reproduce the experimental
spectrum. Care was taken to maintain a reasonable width for
each transition, varying between 0.25 eV for sharp 1s f π*
transitions and 2-4 eV for 1s f σ* and continuum transitions.
These values were chosen to reflect the widths observed in
previous experimental spectra, as well as to reasonably repro-
duce the spectra measured in this work. Transition assignments
were obtained by examining the spatial extent of the molecular
orbital of the final state and by comparison of relative oscillator
strengths and positions. Varying the exchange/correlation func-
tionals and basis sets used was found to change the absolute
value and position of the calculated transitions, but the relative
spacing and intensities were well preserved, requiring only a
constant energetic offset to provide consistent agreement with
the experimental spectra.
To determine the effect of solvent interaction on the NEXAFS
spectra, the process described above was repeated on several
hydrated glycine clusters, each containing a single glycine
molecule surrounded by seven water molecules. No qualitative
difference was observed between the calculated spectra of the
isolated molecule and that of the hydrated cluster at the carbon
and oxygen K edges, and the effect of hydration at the nitrogen
K-edge is discussed below.
3. Results and Discussion
Both subtle and dramatic changes are observed in the
NEXAFS spectra of aqueous glycine upon shifting the pH away
from the isoelectric point (pI). These effects can be rationalized
via a combination of DFT calculations and comparison to the
analogous NEXAFS spectra of solid (thin film) and gas-phase
glycine. To provide a systematic rationalization of these effects,
the next three subsections will address the NEXAFS spectra of
each pertinent atomic region (K-edges of C, N, and O) in order
of increasing energy.
3.1. Carbon K-Edge. Figure 2 shows the carbon K-edge TEY
NEXAFS spectrum of a 0.6 M aqueous glycine solution at pH
) pI ) 5.97 (B), compared to the solid scanning transmission
microscopy (STXM) spectrum (C), and the gas-phase inner-
shell electron energy-loss spectroscopy (ISEELS) spectrum (A)
measured by Gordon et al.1Although the spectra of all three
species are quite similar, there are noticeable differences in the
width and relative intensities of the resonances located between
290 and 300 eV. These resonances have been assigned to
transitions to a combination of σ* and Rydberg states on either
the carboxylic or amine carbons.1,5,6The oscillations apparent
in both the solid and aqueous phase spectra are absent in the
gas phase spectrum. This is most likely due to the fact that the
gas phase spectrum was measured using ISEELS, and the
resulting loss in resolution has broadened these peaks into a
single continuum. Furthermore, the higher-energy resonances
of the solid phase spectrum have also been broadened into a
single feature. As STXM and TEY NEXAFS have very similar
spectral resolution, this broadening is most likely due to
interactions between neighboring glycine molecules within the
crystalline sample. In many crystalline solids, the numerous
Rydberg transitions located around the ionization potential
merge to form a single broad resonance.32As will be demon-
strated below, the density of states within the 295-305 eV range
is quite large, and these resonances can quite easily form a single
continuum band.
Figure 1. Neutral, zwitterionic, cationic, and anionic forms of glycine.
In each case, the minimum-energy geometry was determined for each
charge state. At room temperature, these are the only significantly
populated species for each charge state, as determined by a Boltzmann
weighting of the calculated single point energies of several possible
conformers for each species.
TABLE 1. Geometries Used for StoBe DeMon 2.0 DFT
Calculations Corresponding to Those Shown in Figure 1a
neutral zwitterion
position (Å) position (Å)
atom
XYZ
atom
XYZ
C1
C2
N
O1
O2
H
H
H
H
H
0.1973 -1.1107
0.1973
-1.1183 -1.7546
1.4801
-0.7995
1.3774
0.7765 -1.4414
0.7765 -1.4414 -0.8761
-1.6494 -1.4294
-1.6494 -1.4294 -0.8115
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.8761
C1
C2
N
O1
O2
H
H
H
H
H
0.4140 -1.0024
0.4139
-1.0636 -1.4147
1.5511
-0.7765
-1.2525 -2.4271
0.8850 -1.4175
0.8850 -1.4175 -0.8981
-1.4980 -0.9287
-1.4980 -0.9287 -0.7999
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.8981
0.4322 0.5934
0.9356
1.1507
1.9106
1.1103
1.0763
0.81150.7999
cation anion
position (Å)position (Å)
atom
XYZ
atom
XYZ
C1
C2
N
O1
O2
H
H
H
H
H
H
0.2237 -1.1325
0.2237
-1.2422 -1.5342
1.4669
-0.8340
1.4181
-1.7641 -0.6192
0.7216 -1.5248
0.7216 -1.5248 -0.8928
-1.5078 -2.0756
-1.5078 -2.0756 -0.8325
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.8928
C1
C2
N
O1
O2
H
H
H
H
0.3089 -1.0415
0.3089
-1.0219 -1.7299
1.4734
-0.8400
0.8810 -1.3819
0.8810 -1.3819 -0.8794
-1.5265 -1.3061
-1.5265 -1.3061 -0.7895
0.0000
0.0000
0.0000
0.0000
0.0000
0.8794
0.41360.5486
0.9048
1.0436
1.8877
1.0738
1.1172
0.7895
0.8325
aLowest-energy geometry of each charge species investigated in this
work. See the text for a detailed description of the basis set, excitation
scheme, and exchange/correlation functionals used.
pH Dependence of Electronic Structure of Glycine
J. Phys. Chem. B, Vol. 109, No. 11, 2005 5377

Page 4

As evident in Figure 1, the carbon backbone is expected to
be relatively unaffected by pH, compared to the nitrogen and
oxygen termini. This suggests that any changes in electronic
structure around the carbon atoms will be subtle, and that the
resulting differences in the TEY NEXAFS of the carbon K-edge
will be minor. This prediction is verified in Figure 3, which
depicts the TEY spectra of 0.6 M glycine solutions at pH )
11.86 (A), 5.97 (B), and 1.11 (C). The only noticeable changes
observed as one varies [H+] over 11 decades are a small (0.15
eV) red shift in the first sharp resonance around 287 eV at high
pH and a redistribution in the relative heights and widths of
the broad resonances between 290 and 305 eV. The sharp
resonance has previously been assigned to a 1s f π* transition
localized on the carboxylic carbon,1,5,6whereas the broad
transitions have been assigned to transitions to mixed σ*-
Rydberg states.1,5,6It is somewhat surprising that the shift in
the π* orbital occurs in basic solution, because the carboxyl
terminus is qualitatively identical in the anionic and zwitterionic
species. To rationalize this energy shift, as well as the other
subtle spectral differences observed, the carbon K-edge spectra
for each functional group were calculated for the lowest-energy
conformer of the cation, zwitterion, and anion. The calculated
spectra of the amine (solid) and carboxylic (dotted) carbons are
compared to the corresponding experimental spectra in Figure
4. The stick spectra used to create each spectrum are also
included. From this information, a spectral assignment can be
made for each of the observed species.
A strong peak near 287 eV is observed for all three forms of
glycine. From the calculated spectra and previous assign-
ments1,5,6this can be assigned to the C 1s f πCdO* transition.
The red shift of this transition at high pH is actually overesti-
mated in the calculated spectra (0.375 eV). A smaller red shift
(0.07 eV) is also predicted for the acidic species. Although this
shift is below the experimental resolution of this work, there is
a shift of ∼0.1 eV in the peak position of the C 1s f πCdO*
transition when the acidic and neutral spectra are fit to a
Lorentzian line shape. The relative position of the C 1s f πCd
O* transition is extremely sensitive to the electronegativity of
the neighboring substituents of the carbonyl carbon,33with an
observed red shift of 3.8 eV in poly(vinyl methyl ketone) when
compared to polycarbonate. In light of this sensitivity, the red
shift at high pH can be explained in terms of the decrease in
electronegativity that occurs upon deprotonation. The smaller
shift at low pH can be explained by the breaking of the
degeneracy of the πCOO* orbital that occurs upon protonation
of the carboxylic group. This loss of degeneracy destabilizes
the bonding πCdOorbital, resulting in a stabilization of the πCd
O* system. Another test of this explanation can be realized by
comparing the solid (zwitterionic) and vapor (neutral) spectra
collected by Gordon et al.1The 0.3 eV red shift observed in
the vapor relative to the solid is in agreement with the
incremental changes observed here. That the position of the C
1s f πCdO* transition can be affected by such a minor change
in the electronegativity of the amine carbon is surprising and
highlights the sensitivity of NEXAFS as a probe of chemical
environment and electronic structure.
The broad resonances observed in all three spectra are more
difficult to assign. As evident in Figure 4, the transition density
between 290 and 305 eV is quite large. Because no single
transition in this region appears to dominate the spectrum, each
of the broad resonances can be assigned as a transition to a
variety of σ* and Rydberg states. The first feature around 290
eV appears to be consistently formed by transitions on the amine
carbon, whereas the remaining features appear to have signifi-
cant contribution from both the carboxyl and amine carbons.
This interpretation is in agreement with that of Gordon et al.1
Figure 2. Carbon 1s NEXAFS spectra of glycine. A. Gas-phase signal
measured by inner-shell electron energy loss spectroscopy (ISEELS)
in the dipole regime.1B. Aqueous (0.6 M, pH 5.97) signal measured
by total electron yield (TEY) NEXAFS. C. Thin solid film measured
by scanning transmission X-ray microscopy (STXM).1
Figure 3. Effect of pH on the TEY NEXAFS C K-edge spectra of 0.6
M glycine (aq). The solution pH was 11.86, 5.97, 1.11 for spectra A,
B, and C, respectively. Each spectrum has been background subtracted
and normalized to the excitation flux. Offsets have been included for
clarity and to correct for energy calibration. Special note should be
made of the small (0.15 eV) energy shift observed in the C 1s f π*Cd
Oresonance centered at 289 eV at pH 11.86.
5378 J. Phys. Chem. B, Vol. 109, No. 11, 2005
Messer et al.

Page 5

Ultimately, none of the spectral differences observed for the
carbon K-edge are large enough to alter the interpretation of
the measurements previously made on thin film samples.6
3.2. Nitrogen K-Edge. The nitrogen K-edge TEY NEXAFS
spectrum of aqueous glycine at pH ) pI ) 5.97 (B) is compared
to the solid scanning transmission microscopy (STXM) spectrum
(C) and the gas-phase inner-shell electron energy-loss spectros-
copy (ISEELS) spectrum (A) measured by Gordon et. al.1in
Figure 5. There is little qualitative difference between the solid
and aqueous spectra, but both appear quite distinct from that of
the vapor phase. The main resonance has been red shifted by
1.4 eV in the vapor compared to the bulk. Additionally, a sharp
preedge resonance and a strong shoulder peak are observed at
Figure 4. Experimental C K-edge TEY spectra and corresponding calculated spectra for 0.6 M glycine as a function of pH. A. pH 1.11. B. pH
5.97. C. pH 11.86. From top to bottom, each graph includes the experimental spectrum and calculated spectrum (thick black), as well as the
calculated amine carbon (C1) spectrum (thin gray), and calculated carboxylic carbon (C2) spectrum (thin black). The calculated spectra were created
via a Gaussian convolution of the computed stick spectra, which have been included for reference under each curve. A precise spectral assignment
(1s f π*CdO) is only possible for the sharp resonance observed near 289 eV, as the large spectral density between 290 and 305 eV creates a broad
continuum over this energy range.
Figure 5. Nitrogen 1s NEXAFS spectra of glycine. A. Vapor signal
measured by inner-shell electron energy loss spectroscopy (ISEELS)
in the dipole regime.1B. Aqueous (0.6 M) signal measured by total
electron yield (TEY) NEXAFS. C. Thin solid film measured by
scanning transmission X-ray microscopy (STXM) from Gordon.1
Figure 6. pH dependent TEY NEXAFS N K-edge spectra of 0.6 M
glycine (aq). The solution pH was held at 11.86, 5.97, 1.11 for spectra
A, B, and C, respectively. Each spectrum has been background
subtracted and normalized to the excitation flux. Offsets have been
included for clarity and to correct for energy calibration. No qualitative
spectral differences are observed between the acidic (pH 1.11) and
neutral (pH 5.97) solutions. There is a large (1.3 eV) energy shift in
the main edge for the basic (pH 11.86) solution, as well as the
appearance of two sharp preedge resonances at 405 and 406.3 eV. These
sharp resonances have been assigned to the N 1s f σ* transitions.
pH Dependence of Electronic Structure of Glycine
J. Phys. Chem. B, Vol. 109, No. 11, 2005 5379