An experimental and computational investigation into the gas-phase acidities of tyrosine and phenylalanine: Three structures for deprotonated tyrosine

Article

pubs.acs.org/JPCB

An Experimental and Computational Investigation into the Gas-

Phase Acidities of Tyrosine and Phenylalanine: Three Structures for

Deprotonated Tyrosine

Samantha S. Bokatzian,^†Michele L. Stover, Chelsea E. Plummer, David A. Dixon,*

and Carolyn J. Cassady*

Department of Chemistry, The University of Alabama, Tuscaloosa, Alabama 35487, United States

S

* Supporting Information

ABSTRACT: Using mass spectrometry and correlated molecular orbital

theory, three deprotonated structures were revealed for the amino acid

tyrosine. The structures were distinguished experimentally by ion/molecule

reactions involving proton transfer and trimethylsilyl azide. Gas-phase

acidities from proton transfer reactions and from G3(MP2) calculations

generally agree well. The lowest energy structure, which was only observed

experimentally using electrospray ionization from aprotic solvents, is

deprotonated at the carboxylic acid group and is predicted to be highly

folded. A second unfolded carboxylate structure is several kcal/mol higher in

energy and primarily forms from protic solvents. Protic solvents also yield a

structure deprotonated at the phenolic side chain, which experiments ﬁnd to be intermediate in energy to the two carboxylate

forms. G3(MP2) calculations indicate that the three structures diﬀer in energy by only 2.5 kcal/mol, yet they are readily

distinguished experimentally. Structural abundance ratios are dependent upon experimental conditions, including the solvent and

accumulation time of ions in a hexapole. Under some conditions, carboxylate ions may convert to phenolate ions. For

phenylalanine, which lacks a phenolic group, only one deprotonated structure was observed experimentally when electrosprayed

from protic solvent. This agrees with G3(MP2) calculations that ﬁnd the folded and unfolded carboxylate forms to diﬀer by 0.3

kcal/mol.

sulfur at the side chain and the deprotonated C-terminal

carboxylate group. An infrared spectroscopy study suggested

that the carboxylic acid group was deprotonated in cysteine, but

this still can allow for the shared hydrogen between the two

anionic sites.⁵In the context of a peptide backbone, the

carboxylic acid functionality becomes part of the amide linkage

except at the C-terminus, thus removing a likely deprotonation

site. However, deprotonation of peptides lacking highly acidic

sites has been observed experimentally.^6,7Thus, alternative

deprotonation sites are accessible using common mass

spectrometry ionization techniques such as electrospray

ionization (ESI).

INTRODUCTION

■

The amino acids tyrosine and phenylalanine often have similar

properties because both possess aromatic, hydrophobic side

chains. Their structures (shown in Chart 1) are similar with

Chart 1

Kass and co-workers have studied deprotonated tyrosine

experimentally and theoretically, with a focus on determining if

the deprotonated ion has a carboxylate or phenoxide

structure.^8,9These researchers developed a useful ion/molecule

reaction involving trimethylsilyl azide (TMSN₃) to distinguish

between carboxylate and phenoxide ions.^8,9They employed this

chemical probe to study the eﬀects of the solvent system on

deprotonated tyrosine structures produced by ESI and found

that phenoxide is favored when the solvent includes methanol,

while carboxylate dominates when acetonitrile or acetonitrile/

tyrosine having a 4-hydroxybenzyl group at the side chain and

phenylalanine having a benzyl side chain. Tyrosine possesses

two acidic functionalities capable of deprotonation, the

carboxylic acid group and the phenolic hydroxyl group. In

the gas phase, most amino acids and peptides deprotonate at

the C-terminal carboxylic acid group (if such a group is

present). Exceptions include aspartic acid and glutamic acid,

which have carboxylic acid groups located on the side chain and

can deprotonate readily on either the C-terminus or the side

chain.¹Another exception is cysteine, which has been shown to

deprotonate on the side chain,²⁻⁴with theory³and experiment⁴

indicating that a proton is shared between the deprotonated

Received: October 3, 2014

Published: October 9, 2014

12630

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

water solvents are used.⁸Disparate results from photoelectron

spectroscopy (PES) experiments indicate that the carboxylate

ion is dominant.⁹Kass and co-workers attributed this

inconsistency between their TMSN₃and PES experiments to

diﬀerent ESI source conﬁgurations. The PES experiments,

which found carboxylate ions, used a home-built ESI source in

which ions leaving the skimmer region pass through an ion

guide for 100 ms. The ion/molecule reactions with TMSN₃,

which found primarily phenoxide ions, employed a commercial

ESI source where ions leaving the skimmer are accumulated in

a hexapole trap for 1−5 s prior to their introduction into the

mass analyzer. Using density functional theory (DFT)

calculations (B3LYP/aug-cc-pVDZ), Kass and co-workers

found a carboxylate structure to be energetically favored by

0.2 kcal/mol over a phenoxide structure.⁹This very small

diﬀerence would suggest the existence of both structures in the

gas phase. These researchers interpreted their results as

deprotonated tyrosine undergoing a structural change as the

compound goes from a solution-phase neutral to a gas-phase

anion.⁸

Oomens and co-workers performed infrared multiphoton

dissociation (IRMPD) on several deprotonated amino acids,

including tyrosine.⁵The IRMPD spectra were compared to

calculated spectra at the DFT B3LYP/6-31++G** level of

theory. The amino acids were dissolved in a 80:20 (v/v) ratio

of methanol (CH₃OH) and water (H₂O), which is similar to

the solvent conditions where Tian and Kass⁸observed

increased phenoxide production. The ESI source used by

Oomens and co-workers⁵had a very similar commercial design

to the source used by Kass and co-workers⁹for the TMSN₃

reactions; both sources employed hexapole accumulation times

of several seconds. However, the experimental IRMPD results

show that deprotonated tyrosine is a carboxylate in the gas

phase, with no evidence of phenoxide. We previously

predicted³that tyrosine is deprotonated at the C-terminal

carboxylate site using a variety of correlated molecular orbital

(MO) theory methods up through CCSD(T)/aug-cc-

pVTZ.^10,11Our higher level calculations show that the

carboxylate ion is more stable than the phenoxide ion by ∼2

kcal/mol, which is consistent with the IRMPD experimental

results. Using lower level DFT, Li et al.¹²have found that the

experimental IRMPD spectra of Oomens and co-workers⁵are

best matched to the calculated IR spectra for the carboxylate

ion.

Gas-phase acidity (GA or ΔG_acid) values are an important

tool in understanding structure, reactivity, and fragmentation

behavior of compounds in mass spectrometry. The GA is the

Gibbs free energy change (ΔG) for the reaction AH → A⁻+

H⁺. GAs or deprotonation enthalpies (ΔH_acid) have been

determined for the amino acids by O’Hair and co-workers¹³

and Poutsma and co-workers¹⁴using the kinetic and extended

kinetic methods, respectively, which involve collision-induced

dissociation (CID) of a proton-bound dimer containing the

analyte and a reference compound of known acidity. Poutsma

and co-workers also calculated ΔH_acidfor the common amino

acids using a hybrid DFT approach with B3LYP functional

combinations.¹⁴We have determined the GAs of glutamic acid

and aspartic acid with DFT and molecular orbital (MO)

computational approaches, as well as experimentally using the

thermokinetic method,¹⁵which involves measurement of the

rates of ion/molecule reactions between the deprotonated ion

and reference compounds of known GA.¹In addition, we

predicted the GAs for the 20 common and 5 rare amino acids

using the G3(MP2) method and obtained results consistent

with experiment and with other calculations.³Kass and co-

workers found the GA of tyrosine to be 332.5 1.5 kcal/mol

using equilibrium GA measurements in a dual cell Fourier

transform ion cyclotron resonance mass spectrometer (FT-

ICR).⁹

In our recent study of the gas-phase and aqueous acidities of

the amino acids, tyrosine was studied at eight computational

levels involving both DFT and MO theory.³In all cases, a gas-

phase deprotonated carboxylate structure was predicted to be

more stable on the free energy scale by 1.7−2.7 kcal/mol than a

deprotonated phenoxide structure. However, as discussed

above, gas-phase equilibrium proton transfer reactions and

ion/molecule reactions involving TMSN₃have found the

phenoxide structure to be produced in the greatest abundance

from protic solvents using ESI.^8,9This diﬀerence was the

motivation for the current study. Our goal was to

experimentally ﬁnd the lowest energy, most acidic carboxylate

form of gas-phase deprotonated tyrosine.

EXPERIMENTAL AND COMPUTATIONAL

METHODS

■

Mass Spectrometry. All experiments were performed with

a Bruker Daltonics (Billerica, MA, USA) BioApex 7T FT-ICR

mass spectrometer. The amino acids were in the ^L-stereo-

isomer. Samples were prepared at 60 μM in solvents containing

various ratios of CH₃OH, ultrapure H₂O, acetonitrile

(CH₃CN), acetone, dimethyl sulfoxide (DMSO), tetrahydro-

furan (THF), and N-dimethylformamide (DMF). All of the

organic solvents were HPLC or LC-MS grade, except DMSO,

which was ACS grade. For some solutions, 1% (by volume) of

ammonium hydroxide (NH₄OH) was added to promote

deprotonation. For experiments in which it was important to

have a low water content, the solvents were also dried with 3A

pore size molecular sieves.

Analyte solutions were introduced into an Apollo API source

using a syringe pump set to deliver ∼90 μL/h. Electrospray

ionization (ESI) employed dried air as a heated (225 °C)

counter and parallel current drying gas. (The air was dried with

a Labclear (Oakland, CA, USA) reﬁllable gas ﬁlter employing

RK-400 molecular sieves with a 13× pore size; the sieves also

contained Drierite as a color changing indicator to denote the

presence of water. The sieves were dried for at least 15 h prior

to experimental use in an oven at 200 °C.) The ESI needle was

grounded, while the capillary entrance and end plate were at a

potential of 3.5−4.0 kV for negative ion mode analysis. Unless

otherwise noted, ions produced by ESI were accumulated in a

hexapole for 600−700 ms before being transported to the

reaction cell by electrostatic focusing. The time used to

transport ions into the cell was 10 ms.

Deprotonated ions, [M − H]⁻, were isolated using correlated

frequency ion ejection techniques.¹⁶Because the ions are

exposed to constant pressures of neutral reactants during this

isolation period, the time involved in ion isolation was kept to a

minimum and was never more than 50 ms. The isolated

precursor ions were then allowed to react with a reference

compound introduced to the ICR cell through a leak valve at a

constant pressure. Each of the ions selected for study were

reacted with a series of reference compounds of known GA.¹⁷

The reference compound pressures were measured using an ion

gauge that was calibrated with the proton transfer reaction

between protonated glycine and N,N-dimethylformamide,

which has an experimental rate constant of 8.19 ( 1.07) ×

12631

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

Table 1. G3(MP2) Gas-Phase Acidities of the Carboxylic Acids and Phenols in kcal/mol

ΔH₂₉₈

G3(MP2)

344.7

ΔH₂₉₈gas

ΔG₂₉₈

G3(MP2)

337.3

ΔG₂₉₈gas

a

compound

formic acid

acid R group

expt

H

345.2 2.9⁵⁴

345.4 2.2⁷⁹

345.3 2.2⁸⁰

346.2 1.2⁸¹

340.1 4.6⁴⁷

348.6 2.1⁵⁴

348.7 2.2⁷⁹

348.1 2.2⁸²

348.2 1.4⁵⁵

343.2 0.7⁴⁷

347.4 2.9⁵⁴

347.4 2.2⁸⁰

346.0 2.1⁸⁰

345.0 2.1⁸³

344.6 2.1⁸⁰

346.5 2.2⁵⁴

346.5 2.2⁸⁰

346.8 2.0⁸⁴

346.7 2.1⁸²

345.5 2.1⁸⁰

345.1 2.1⁸²

344.8 2.1⁸⁰

346.1 2.4⁵⁹

346.2 2.1⁸⁰

338.2 2.0⁵⁴

338.4 2.0⁷⁹

338.3 2.0⁸⁰

339.2 1.5⁸¹

b

344.5

337.0

acetic acid

CH₃

348.4

340.3

b

348.3

341.5

341.5 2.0⁵⁴

341.7 2.0⁷⁹

341.1 2.0⁸²

propanoic acid

CH₂CH₃

347.9

340.0

340.3 2.0⁵⁴

340.4 2.0⁸⁰

339.0 2.0⁸⁰

338.0 2.0⁸³

337.6 2.0⁸⁰

339.5 2.0⁵⁴

339.5 2.0⁸⁰

isobutyric acid

CH(CH₃)₂

C(CH₃)₃

346.7

345.3

338.9

337.7

trimethylacetic acid

butanoic acid

CH₂CH₂CH₃

347.3

339.4

isovaleric acid

CH₂CH(CH₃)₂

CH₂C(CH₃)₃

345.9

344.8

347.1

338.5

337.4

339.1

339.7 2.0⁸²

338.5 2.0⁸⁰

338.1 2.0⁸²

337.8 2.0⁸⁰

339.1 2.3⁵⁹

339.2 2.0⁸⁰

tert-butyl-acetic acid

pentanoic acid

CH₂CH₂CH₂CH₃

isohexanoic acid

tert-butyl-propanoic acid

phenol

CH₂CH₂CH(CH₃)₂

CH₂CH₂C(CH₃)₃

346.5

346.2

349.5

338.7

337.8

342.1

348.3 1.7⁸⁵

348.0 1.0⁵⁵

350.3 2.3⁵⁴

349.1 2.1⁸⁶

350.8 0.9⁸⁷

350.4 3.1⁸⁸

347.5 1.9⁸⁹

350.7 2.9⁹⁰

349.5 2.2⁷⁹

350.7 2.3⁹⁰

349.5 2.1⁷⁹

351.6 2.3⁹⁰

350.2 2.1⁸⁶

350.2 2.1⁷⁹

341.5 1.8⁸⁵

343.4 2.0⁵⁴

342.3 2.0⁸⁶

340.8 1.9⁸⁹

o-cresol

m-cresol

p-cresol

348.6

349.9

350.7

341.4

342.7

342.9

342.7 2.0⁹⁰

342.0 2.0⁷⁹

343.8 2.0⁹⁰

342.7 2.0⁷⁹

344.7 2.0⁹⁰

343.4 2.0⁸⁶

343.4 2.0⁷⁹

a

b

The superscripted numbers provide the reference for each literature value. CCSD(T) values.

10⁻¹⁰cm³/molecules·s.¹⁸Pressures were corrected for reactant

gas ionization eﬃciency,¹⁹which involved polarizabilities

calculated by atomic hybrid parameter procedures.²⁰Reference

compound pressures were in the range (1−10) × 10⁻⁸mbar.

For the reactions of each precursor ion with each reference

compound, at least one pressure used was from the upper half

of this range and at least one pressure was from the lower half

of this range; the higher and lower pressures diﬀered by at least

a factor of 4 (×4).

Reaction rate constants, k_exp, were obtained by observing the

pseudo-ﬁrst-order decay in reactant ion intensity as a function

of trapping time. In cases where deprotonation was in

competition with proton-bound dimer formation, k_expwas

obtained by ﬁtting the experimental ion intensity data as

discussed previously.²¹For experiments in which nonlinear

pseudo-ﬁrst-order kinetics plots (bimodal plots) indicated the

presence of two ion structures reacting at two diﬀerent rates,

the data was ﬁt to the sum of two exponential decays using the

program Sigma Plot by Systat Software Inc. (San Jose, CA,

USA). The fraction that each exponential contributes to the ﬁt

directly relates to the relative abundance of that ion structure,

while the slope of each exponential decay is used to calculate

the reaction rate constant (in the same manner that this

information is used to calculate rate constants from a unimodal

kinetics ﬁt). This procedure has been used in the past to obtain

rate constants²²⁻²⁶and gas-phase basicities²⁷⁻²⁹for systems

containing two or three ion structures reacting at diﬀerent rates.

The ratio of the experimental rate constant to the thermal

capture rate constant^30,31yields a reaction eﬃciency (RE). A

RE of 0.269 was used as a break point, where a reaction is

considered to become exoergic and a GA value is assigned. This

selection of 0.269 comes from the work of Bouchoux and co-

workers,^15,32−34which has been termed the “thermokinetic

method”. We have found that this method provides excellent

12632

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

agreement between experimental and theoretical GAs for

several amino acids and small peptides.^1,6

acid resulted in an increased acidity of 0.5 kcal/mol. These

trends follow for the other carboxylic acids as well. Carboxylic

acids with two methyl groups are slightly more acidic and range

from 338.5 to 338.9 kcal/mol, and those with three methyl

groups are even more acidic, ranging from 337.4 to 337.8 kcal/

mol.

Computational Methods. The calculations were per-

formed at the DFT and correlated MO theory levels with the

program Gaussian 09.³⁵The geometries were initially

optimized at the DFT level with the B3LYP exchange-

correlation functional^36,37and the DZVP2 basis set.³⁸Vibra-

tional frequencies were calculated to show that the structures

were minima. A range of structures were optimized to

determine the most stable conformers chosen by sampling

many conformations with and without hydrogen bonds. A

substantial number of low energy conformers were found for

the neutrals as discussed below. In our previous work on

predicting the GAs of amino acids^1,3,39and organic acids,³⁹we

showed that the high level G3(MP2) correlated MO method⁴⁰

gave agreement for the acidities with the experimental values to

within about 1 kcal/mol and also agreement with higher level

CCSD(T) calculations extrapolated to the complete basis set

limit with additional corrections.⁴¹⁻⁴⁵G3(MP2) has an

additional advantage over DFT methods in terms of reliable

predictions for these types of compounds because the

correlated MO methods in G3(MP2) perform better in the

prediction of hydrogen bond energies as well as steric

nonbonded interactions than do most widely used DFT

exchange-correlation functionals. Thermal corrections to the

enthalpies and the free energies were calculated in the scaled

harmonic oscillator, rigid approximation⁴⁶using the geometries

and frequencies obtained in the G3(MP2) calculations

(Hartree−Fock/6-31G* level).

Organic Acid and Alcohol Benchmarks. As a further

benchmark of the computational methods in use, the GAs have

been predicted for 11 carboxylic acids (formic, acetic,

propanoic, isobutyric, trimethylacetic, butanoic, isovaleric, tert-

butylacetic, pentanoic, isohexanoic, and tert-butylacetic) and 4

phenols (phenol and ortho-, meta-, and para-cresol). The

optimized structures of the neutrals and resulting anions of the

11 carboxylic acids are shown in the Supporting Information.

Conformational searches were done, and the lowest energy

conformer of the neutral and anion was used for the acidity

calculations. In all cases, except for isobutyric acid, the

orientation of the methyl groups in the neutrals and their

corresponding anions were identical. In isobutyric acid, a

rotation of the two methyl groups occurred to produce the

most stable anion.

The optimized structures for the neutrals and anions of

phenol and the three cresols are given as Supporting

Information. The calculated values of the phenol and cresol

acidities are all within the experimental error bars, as shown in

Table 1. The experimental results show that the acidity of the

cresols decreases from ortho to meta to para with the acidity of

phenol falling between o-cresol and m-cresol, giving the order o-

cresol > phenol > m-cresol > p-cresol. The diﬀerence in

acidities is only 2 kcal/mol from o-cresol to p-cresol. As the

acidity diﬀerence is small, further calculations were performed

to demonstrate that p-cresol was not lower in energy at a

diﬀerent computational level. The energy diﬀerences between

the ortho, meta, and para anions were calculated at diﬀerent Gx

computational levels,⁵⁰⁻⁵³as shown in the Supporting

Information. o-Cresol was predicted to be ∼1 kcal/mol more

acidic than m-cresol with all methods except for G3B3 and

G3MP2B3 where o-cresol was predicted to be more acidic by

∼2 kcal/mol. o-Cresol was predicted to be more acidic than p-

cresol by 1.5−2 kcal/mol.

RESULTS AND DISCUSSION

■

Calculated GAs and Structures for Tyrosine, Phenyl-

alanine, and 4-(4-Hydroxyphenyl)-2-butanone (HPB).

For deprotonated tyrosine, we have previously predicted at

the G3(MP2) level that the carboxylate anion is more stable

than the phenoxide by 2.5 kcal/mol.³The structure of the

carboxylate is folded to maximize hydrogen bonding. The

question exists as to whether there are additional higher energy

carboxylate conformers with a similar elongated structure to

that of the phenoxide. Figure 1 shows the optimized structures

of the lowest energy conformer of the two amino acids and

HPB and low energy conformers for the deprotonated species.

Unless noted, all energies refer to free energies at 298 K. The

multiple low energy conformers of the neutral amino acids are

given in the Supporting Information, and in all cases, the lowest

energy folded and unfolded structures diﬀer by less than 0.5

kcal/mol. Table 2 shows the experimental and G3(MP2)

calculated GAs for these molecules.

Excellent agreement is found with the available experimental

ΔH_acidand GA values of the organic acids except for the results

from Muftakhov et al.⁴⁷(Table 1). Muftakhov et al.

underestimate the deprotonation enthalpies of formic and

acetic acid by ∼5 kcal/mol in both cases. For formic acid, the

reaction enthalpy was also calculated at a composite CCSD-

(T)/CBS level⁴¹⁻⁴⁵to further benchmark the ability of

G3(MP2) to calculate the acidities for these molecules. The

components for the CCSD(T) atomization energies are given

in the Supporting Information. The CCSD(T) results for

formic and acetic acid³⁹(Table 1) are in excellent agreement

with the GAs calculated at the G3(MP2) level and the

experimental values. The CCSD(T) calculations¹⁰were done

with the augmented correlation consistent basis sets¹¹and the

MOLPRO⁴⁸and NWChem⁴⁹program systems.

There are 22 conformers for tyrosine with energies within 2

kcal/mol of the lowest energy structure. Weak hydrogen

bonding between the amine and two carbons on the phenolic

ring plays an important role in the neutral, leading to a folded

structure. For the tyrosine anion, we studied 10 diﬀerent

structures. Three low energy structures (Figure 1a) were found

that diﬀer by only 2.5 kcal/mol, two carboxylate conformers

and one phenoxide isomer, in contrast to the large number of

low-lying structures in the neutral. A folded structure is

predicted to be the lowest energy structure for the carboxylate

−

anion with ring C−H’s interacting with the CO₂group. The

carboxylate anion has a higher energy unfolded conformer 1.0

kcal/mol less stable than the folded carboxylate conformer. The

lowest energy conformer for the phenoxide anion is unfolded, is

2.5 kcal/mol less stable than the lowest energy folded

carboxylate conformer, and is 1.5 kcal/mol less stable than

the unfolded carboxylate conformer on the free energy scale at

298 K. In terms of ΔH₂₉₈(gas), the energy diﬀerence between

the unfolded anions is smaller, only 0.3 kcal/mol.

The results show that the carbon chain length has only a very

small eﬀect on the acidity. Upon comparison of the acidities of

butanoic and isobutyric acids, which both have four carbons in

the chain, we found the additional methyl group in isobutyric

12633

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

anion is folded and 2.3 kcal/mol more stable than the unfolded

ion.

Experimental GAs and Structures Produced from a

Protic Solvent System. Our experimental studies began with

proton transfer ion/molecule reactions to bracket the GAs of

tyrosine, phenylalanine, and HPB. This work was performed

using a protic solvent mixture of 49.5:49.5:1 (v/v/v)

CH₃OH:H₂O:NH₄OH, which is a routine ESI solvent system

that often generates intense negative ions. Table 3 shows the

reference compounds used in this study, their GAs, and the

measured reaction eﬃciencies. From these results, experimental

GAs were assigned. Our experimental and computational GA

values are summarized in Table 2, along with literature values

from other studies on these compounds.

The least acidic compound studied was HPB, which has an

experimental GA of 339.6 3.0 kcal/mol. (A higher GA value

indicates a less acidic compound.) This agrees with our

G3(MP2) calculated value of 337.8 kcal/mol, and the

computational result is well within the experimental error

bars. HPB was chosen for comparison to tyrosine and

phenylalanine because this compound has no carboxylic acid

group and, therefore, must deprotonate at its phenolic hydroxyl

group.

Phenylalanine, which contains no phenolic side chain but has

a carboxylic acid group at the C-terminus, is more acidic than

HPB and has an experimental GA of 332.5

2.2 kcal/mol,

which agrees within experimental error with our G3(MP2)

value of 330.7 kcal/mol. For HPB and phenylalanine, all

experimental kinetics plots were linear, suggesting the presence

of only one major deprotonated ion structure. If multiple

structures exist, the experiments indicate that they are very

similar in energy; for example, the 0.3 kcal/mol energy

diﬀerence found by G3(MP2) calculations between the folded

and unfolded conformers of deprotonated phenylalanine would

be diﬃcult to distinguish experimentally. There is reasonable

agreement of both the calculated and experimental GAs with

the experimental GA for phenylalanine of 329.6 3.0 kcal/mol

previously reported by O’Hair et al.¹³The fact that HPB is less

acidic than phenylalanine is consistent with phenols generally

being less acidic than carboxylic acids. For example, the GA of

phenol is 343.4 2.0 kcal/mol, while the GA of acetic acid is

341.5 2.0.⁵⁴Recently calculated GAs by Angel and Ervin⁵⁵

based on experimental enthalpies for loss of a proton are 341.5

1.0 and 339.9 1.7 for phenol and acetic acid, respectively,

and are consistent with the above values. In both cases, the

error bars overlap but the experimental results and the

theoretical results in Table 1 are all consistent in the ordering

between acetic acid and phenol.

Figure 1. G3(MP2) results showing the structures of the most stable

neutral acid and the lowest energy anions resulting from proton loss.

Energies are in kcal/mol. Hydrogen bond distances are in Å. The

yellow arrows show the deprotonation site. (a) Tyrosine, (b)

phenylalanine, and (c) HPB.

As shown in Table 2, we have previously used a variety of Gx

methods^51,52and CCSD(T) theory to predict the energy

diﬀerence between the folded carboxylate and phenoxide sites

of tyrosine. We used these same methods to calculate the

relative energy values for the unfolded carboxylate anion. In all

cases, the folded carboxylate anion is predicted to be 1.7−2.7

kcal/mol more stable than the phenoxide and the unfolded

carboxylate is between the folded carboxylate and the unfolded

phenoxide. Higher energy deprotonation sites include the

aromatic ring, the α carbon, and the amine group, and these are

given in the Supporting Information for the amino acids.

Phenylalanine has 13 low energy conformers within 2 kcal/

mol of the lowest energy folded structure. Deprotonation of

phenylalanine (Figure 1b) forms two carboxylate structures that

have G3(MP2) acidities diﬀering by 0.3 kcal/mol with the

lowest energy conformer being folded.

Tyrosine is an interesting case because there are two

potential sites of deprotonation. Using the

CH₃OH:H₂O:NH₄OH solvent system, two ion populations

were observed in the reactions with ethyl cyanoacetate (GA =

333.6

2.0 kcal/mol¹⁷) and 4-amino-2,3,5,6-tetraﬂuoropyr-

idine (ATFP, GA = 332.8 2.0 kcal/mol¹⁷). To illustrate this

bimodal reactivity, Figure 2 gives a semilogarithmic plot of the

intensity of deprotonated tyrosine, [Tyr−H]⁻, as a function of

reaction time with ethyl cyanoacetate. The experimental data

(black circles) are an excellent ﬁt to an equation involving the

sum of two exponentials (black line). From the reaction

eﬃciency data of Table 3, the less acidic deprotonated tyrosine

structure accounts for ∼27% of the ions and has a GA of 333.5

2.4 kcal/mol, while the more acidic structure accounts for

∼73% of the ions and has a GA of 332.4 2.2 kcal/mol. This is

Neutral HPB has two low energy conformers, and the lowest

energy structure is unfolded. The lowest energy phenoxide

12634

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

Table 2. Experimental, Theoretical, and Literature Gas-Phase Acidities in kcal/mol for Tyrosine, Phenylalanine, and HPB

experimental

theoretical

experimental

method

deprotonation

site

conformation ΔG₂₉₈gas (GA)

ΔH₂₉₈gas

computational method

G3(MP2)

ΔG₂₉₈gas ΔH₂₉₈gas

Tyrosine

a

present work

thermokinetic

bracketing

carboxylate

phenolate

carboxylate

phenolate

carboxylate

phenolate

carboxylate

phenolate

carboxylate

phenolate

carboxylate

phenolate

carboxylate

phenolate

−

unfolded

folded

333.5 2.4

332.4 2.2

324.7 3.6

−

331.4

332.9

330.4

331.1

332.6

330.5

331.2

332.9

330.2

331.2

332.3

330.6

331.3

332.3

330.6

−

340.2

340.5

338.3

339.9

340.3

338.3

340.0

340.5

338.1

339.6

340.1

338.3

339.7

340.1

338.2

339.5

338.3

339.9

339.0

339.2

−

b

−

unfolded

folded

G3B3

b

unfolded

folded

G3

b

unfolded

folded

G4

b

unfolded

folded

CCSD(T)/aT//MP2/aT

b

c

Kass

equilibrium

332.5 1.5

340.7 1.5

B3LYP

G3B3

−

d

Poutsma

extended kinetic

kinetic

337.7 2.6

B3LYP/6-31+G*

−

e

O’Hair

−

329.5 3.0

Phenylalanine

332.5 2.2

336.4 3.1

−

present work

thermokinetic

bracketing

carboxylate

folded

−

G3(MP2)

330.7

331.0

−

338.5

339.6

338.7

−

unfolded

e

Poutsma

extended kinetic

kinetic

−

338.9 4.3

336.5 3.1

B3LYP/6-31+G*

−

e

O’Hair

329.6 3.0

−

4-(4-Hydroxyphenyl)-2-butanone (HPB)

present work

thermokinetic

bracketing

phenolate

folded

339.6 3.0

−

G3(MP2)

337.8

340.1

344.0

347.5

phenolate

unfolded

a

b

c

d

e

“−” indicates that no value was reported. From ref 4. From ref 9. From ref 14. From ref 13.

the ﬁrst time that experimental GAs have been separately

obtained for the diﬀerent deprotonated tyrosine structures.

Using an equilibrium method, Kass and co-workers reported a

single GA value of tyrosine of 332.5 1.5 kcal/mol,⁹which is

in excellent agreement with our GA value relating to the more

acidic structure. Using the kinetic method of CID on a proton

bound dimer, O’Hair et al. obtained a value of 329.5 3.0 kcal/

mol,¹³which agrees to within experimental error. O’Hair’s value

is in excellent agreement with our calculated G3(MP2) value

for the folded carboxylate.

Taken together, the proton transfer and the TMSN₃

reactivity data indicate that the more abundant and more

acidic deprotonated tyrosine species (∼70% of the ions) has a

phenoxide structure and yields an experimental GA of 332.4

2.2 kcal/mol. This corresponds to structure C of Figure 1a,

unfolded phenoxide, which has a calculated GA of 332.9 kcal/

mol. The less abundant and less acidic deprotonated tyrosine

species (∼30%) has a carboxylate structure and an experimental

of GA of 333.5 2.4 kcal/mol. This corresponds to structure B

of Figure 1a, unfolded carboxylate, which has a calculated GA of

331.4 kcal/mol. These structures were generated from a protic

solvent system of 49.5:49.5:1 (v/v/v) CH₃OH:H₂O:NH₄OH.

Kass and co-workers found this same phenoxide/carboxylate

ratio of 70:30 when tyrosine ions were produced from a

solution of 75:25 (v/v) CH₃OH:H₂O.^8,9Thus, our results and

those of Kass both show that the deprotonated phenoxide ion

is the dominant gas-phase species when tyrosine is electro-

sprayed from a protic solvent system.

The proton transfer reaction data clearly shows two

structures for deprotonated tyrosine but does not distinguish

between carboxylate and phenoxide ions. To identify these

structures, we employed the ion/molecule reaction with

TMSN₃developed by Kass and co-workers.^8,9Using a series

of model compounds to characterize reactivity, they found that

deprotonated tyrosine with the carboxylate structure reacts with

−

TMSN₃to form the azide ion, N₃(m/z 42), and a neutral

tyrosine trimethylsilyl ester (TyrOTMS); in contrast, the

phenoxide structure of deprotonated tyrosine silyates at the

phenol group, producing TMSOTyrCO₂(m/z 252) and the

neutral HN₃. In our work, for ion generation from a protic

CH₃OH:H₂O:NH₄OH solvent system, the TMSN₃reaction

showed that the phenoxide structure accounts for 67( 7)% of

the ions, with the remaining 33( 7)% of the ions being

carboxylate.

Our experiments found the GA involving the phenoxide

structure to be 0.5 kcal/mol lower than the GA with the

carboxylate structure. This diﬀers from the G3(MP2) ordering,

which predicts the unfolded carboxylate structure to yield a

lower GA by 1.5 kcal/mol. There are several possible

experimental reasons for this deviation. First, the proton

transfer reactions involving the two deprotonated tyrosine

reactant ion structures may have diﬀerent activation barrier

−

12635

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

Table 3. Reaction Eﬃciencies for the Proton Transfer Reactions of Tyrosine, Phenylalanine, and HPB with Reference

Compounds

average reaction eﬃciency ( standard deviation)

a

b

c

d

reference compound

chloroform

GA (kcal/mol)

HPB

tyrosine (protic solvent)

tyrosine (aprotic solvent)

phenylalanine

e

349.9

346.0

342.3

341.1

2

0.01 0.00

−

4-triﬂuoromethyl aniline

phenol

−

acetic acid

0.07 0.01

f

BREAK

formic acid

339.2

338.5

337.6

336.2

333.6

2

0.34 0.01

0.41 0.11

0.41 0.06

0.55 0.04

−

NR

−

NR

−

NR

isovaleric acid

g

trimethylacetic acid

p-chlorophenol

ethyl cyanoacetate

NR

h

0.23 0.06 (24 9%)

0.03 0.04 (76 9%)

BREAK

0.028 0.019 (56 7%)

NR (44 7%)

0.007 0.001

4-amino-2,3,5,6-tetraﬂuoropyridine

332.8

2

−

0.53 0.25 (30 8%)

0.07 0.01 (70 8%)

BREAK

0.033 0.011 (61 11%)

0.005 0.004 (39 11%)

BREAK

0.02 0.00

BREAK

3-triﬂuoromethyl phenol

332.4

326.9

2

−

0.29 0.02

0.29 0.03 (58 13%)

0.007 0.002 (42 13%)

0.38 0.04 (59 3%)

0.13 0.04 (41 3%)

BREAK

0.36 0.02

3,3,3-triﬂuoropropionic acid

0.75 0.09

0.79 0.08

diﬂuoroacetic acid

pentaﬂuorophenol

323.8

320.8

2

−

0.45 0.10 (46 13%)

0.33 0.08 (54 13%)

1.2 0.27 (58 15%)

−

0.51 0.03 (42 15%)

a

b

c

All reference compound GAs were obtained from ref 17. HPB stands for 4-(4-hydroxyphenyl)-2-butanone. Protic solvent system of 49.5:49.5:1

d

e

f

(v/v/v) CH₃OH:H₂O:NH₄OH. Aprotic solvent system of 99:1 (v/v) CH₃CN:H₂O. “−” indicates that no reaction was performed. “BREAK”

indicates the point where experimental GA was assigned. “NR” indicates that no reaction was observed. Two reaction eﬃciencies and the relative

g

h

abundances of each are listed for reactions in which bimodal kinetics plots indicate the existence of two ion populations reacting at two diﬀerent

rates.

proton bound dimer involving deprotonated tyrosine and an

acidic reference compound, [(Tyr−H)⁻·H⁺·(A−H)⁻]. This is

essentially the same type of complex that is used in the kinetic

method of gas-phase basicity (GB) and GA determinations,

which involves collision-induced dissociation of the dimer.^59,60

For several small organic acids, deviations from expected GAs

or GBs using the kinetic method have been attributed to the

dimer adapting a slightly higher energy structure in situations

where double hydrogen bonding exists within the dimer or

where the analyte conformation preferred in the dimer diﬀers

from that of its monomer form.⁶¹⁻⁶³For example, Fournier et

al.⁶²found that the amino acid glutamic acid has bimodal

dissociation plots during a GA study using the kinetic method.

Using calculations at the G3(MP2) and OLYP/aug-cc-pVTZ

levels, they concluded that glutamic acid could exist within the

dimer in both zwitterionic and non-zwitterionic forms. Finally,

steric hindrance of the reactive site, either from bulky

substituents or folded conformations with hydrogen bonding,

can lessen the ability of a neutral to access an ion’s reactive site,

which lowers the reaction eﬃciency and results in a lower GA

assignment.^21,64,65

As a check of the computational method, other approaches

were used to predict the energy diﬀerence between the two

carboxylate conformers and phenoxide isomer, as shown in

Table 2. The predicted energy diﬀerences are essentially

independent of the computational method. Geometries were

optimized at the MP2/aug-cc-pVTZ level with diﬀuse functions

and higher order polarization, in contrast to the G3 geometries.

Figure 2. Reactant loss curve for the reaction of deprotonated tyrosine

with ethyl cyanoacetate, which is present at a constant pressure of 8.9

×

10^{− 8}mbar. The solvent is 49.5:49.5:1 (v/v/v)

CH₃OH:H₂O:NH₄OH. The logarithm of [Tyr−H]⁻intensity is

plotted as a function of reaction time. The experimental data points

(black circles) are ﬁtted to an equation involving the sum of two

exponential decays.

heights, which could lead to a reaction being under kinetic

control as opposed to thermodynamic control. This issue is

known to aﬀect proton transfer studies of several organic

molecules.⁵⁶⁻⁵⁸In addition, the reaction intermediate is a

12636

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

Table 4. Reactions of Trimethylsilyl Azide with Deprotonated Tyrosine Electrosprayed from Various Solvents

a

solvent A

solvent B

solvent C

name

ratios of TMSN₃product ions

% phenoxide % carboxylate

62 38

name

H₂O

%

name

%

1

%

100

99

H₂O

NH₄OH

CH₃OH

CH₃CN

CH₃OH

H₂O

68

61

67

65

71

56

63

67

63

70

53

49

3

32

39

H₂O

50

H₂O

49.5

75

74.5

49.5

50

49.5

50

2

NH₄OH

1

33

H₂O

25

35

H₂O

24.5

49.5

50

NH₄OH

1

29

H₂O

44

H₂O

37

CH₃CN

acetone

DMF

49.5

50

NH₄OH

1

33

37

97

NH₄OH

1

30

98

H₂O

1

47

98

H₂O

2

51

98

CH₃OH

NH₄OH

CH₃OH

H₂O

1

NH₄OH

1

97

99

1

35

1

65

99

1

99

1

0

100

0

100

45

0

THF

45

DMSO

10

0

a

All solvent compositions are % by volume.

Relative energies were calculated at the CCSD(T)/aug-cc-

pVTZ level using optimized MP2/aug-cc-pVTZ geometries

with thermal corrections, entropies, and zero point energies

calculated at the MP2/aug-cc-pVDZ level. The CCSD(T)

results are good to at least 1 kcal/mol for the relative energies.

Eﬀects of the Solvent on Deprotonated Tyrosine: A

New Carboxylate Structure. Because the experiments with

the CH₃OH:H₂O:NH₄OH solvent system did not result in our

ﬁnding the low energy carboxylate structure predicted by

theory, additional experiments were performed. Kass and co-

workers had reported that the identity of the solvent used in the

ESI experiment has a dramatic eﬀect on the phenoxide/

carboxylate ratio for deprotonated tyrosine.^8,9They found that

protic solvent systems containing CH₃OH favored phenoxide

formation, while aprotic solvent systems containing CH₃CN

while we found that H₂O and CH₃OH have a similar large

eﬀect on the ratio. For example, as our data in Table 4 indicate,

ESI from a solution of 50% CH₃CN and 50% of either CH₃OH

or H₂O both result in 63% phenoxide ions and 37% carboxylate

ions.

Since aprotic solvents yield almost exclusively carboxylate

ions, we performed proton transfer reactions on deprotonated

tyrosine produced by ESI from a solvent of 99:1 CH₃CN:H₂O.

The mixture included 1% H₂O solution because the presence of

a trace of water greatly increased the solubility of tyrosine;

however, the water content was tightly controlled and the LC-

MS grade CH₃CN was dried with molecular sieves to ensure

that only the 1% water that we added was present. Reactions

with TMSN₃were performed multiple times (including in close

time proximity to the proton transfer reactions) and always

showed that this solvent system forms 99−100% carboxylate

ions and 1 to 0% phenoxide ions.

The reaction eﬃciencies for proton transfer reactions of

tyrosine anions produced from this aprotic solvent are given in

Table 3. For all six reference compounds, two deprotonated

tyrosine ion populations reacting at diﬀerent rates were

observed. To illustrate this, Figure 3 shows the semilogarithmic

plot of the intensity of [Tyr−H]⁻as a function of reaction time

with the reference compound was ATFP. The plot indicates

that ∼71% of the ions are reacting with ATFP (albeit slowly),

while the remaining ions are almost nonreactive. Using the data

of Table 3, the faster reacting (less acidic) species was found to

account for 61( 11)% of the ions and to involve a GA of 333.5

2.4 kcal/mol. The slower reacting (more acidic) species

includes 39( 11)% of the ions and yields a GA of 324.7 3.6

kcal/mol. Because there is ∼9 kcal/mol diﬀerence in acidities,

these two species are readily distinguished by ion/molecule

reactions.

favored the carboxylate structure. In addition, Schroder et al.⁶⁶

̈

found that the phenoxide/carboxylate ratio for deprotonated p-

hydroxybenzoic acid depends on the solvent, pH, and

concentration of the solution being electrosprayed.

Using the TMSN₃ion/molecule reaction to distinguish

between phenoxide and carboxylate ions, we electrosprayed

tyrosine from a variety of solvents. These included seven

solvent systems used by Tian and Kass,⁸plus 13 additional

solvent systems. The results of these experiments are shown in

Table 4. Exclusively (100%) carboxylate ions were found for

the entirely aprotic solvents of pure CH₃CN, pure acetone, and

a mixture of DMF, THF, and DMSO. An addition of 1%

CH₃OH or H₂O to aprotic CH₃CN has almost no eﬀect,

yielding 99−100% carboxylate ions. However, 1% addition of

the base NH₄OH results in 35% of the deprotonated tyrosine

ions having a phenoxide structure. Addition of 2% CH₃OH or

H₂O also has a noteworthy eﬀect, causing about 50−70% of the

ions to be phenoxide. In general, phenoxide ions appear in

abundance when NH₄OH or the protic solvents CH₃OH or

H₂O are used. Our results generally agree with those of Tian

and Kass,⁸although they concluded that CH₃OH greatly aﬀects

the phenoxide/carboxylate ion ratio but H₂O has no eﬀect,

This most acidic structure for deprotonated tyrosine, yielding

an experimental GA of 324.7 3.6 kcal/mol, was not produced

from the protic solvent system. This ion only forms in aprotic

solvent, and the TMSN₃reactions indicate that it has a

12637

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

acidic site of the non-zwitterion of tyrosine, which is the

carboxylic acid group (pK_a= 6.15 for carboxylate deprotonation

and pK_a= 10.37 for phenolate deprotonion³).

To test the premise that the neutral form of tyrosine diﬀers

in protic and aprotic solvents (and that this might account for

diﬀerent gas-phase ion structures), attempts were made to

study the solution-phase structure(s) of tyrosine using FT-IR,

FT-Raman, and NMR. Solutions saturated with tyrosine were

analyzed in solvent systems ranging from 100% aprotic to 100%

protic with 0−5% acid or base additive. Unfortunately, we were

not able to obtain any usable data with these spectroscopic

techniques because of the low solubility of tyrosine in all

solvents tested (including water⁶⁹). The fact that tyrosine can

be studied by mass spectrometry in these solvents is testimony

to the fact that mass spectrometry can analyze very dilute

solutions.

The second, faster reacting tyrosine anion population

generated from aprotic solvent is also interesting. Reactions

with TMSN₃indicate that the aprotic solvent produces 99%

carboxylate ions, but the experimental GA from this second ion

population exactly matches the experimental GA relating to the

phenoxide ion found in protic solvent. This could mean that

the less acidic (more elongated) carboxylate structure is

converting to the phenoxide due to interactions with the

reference compounds. In particular, AFTP is only ∼1 kcal/mol

more acidic than deprotonation of tyrosine to generate the

unfolded carboxylate. During collisions of AFTP neutrals with

tyrosine anions, a proton bound dimer intermediate, [(Tyr−

H)⁻·H⁺·(AFTP−H)⁻], could form. This dimer might dissociate

with conversion of a tyrosine ion from carboxylate to

phenoxide. Such a process would not yield a change in mass

and, therefore, would not have been apparent in the mass

spectra.

Additional support for a conversion between structures is

provided by the ATFP reaction data for tyrosine ions produced

from protic solvent. As seen in Table 3, for tyrosine anions

produced from protic solvent reacting with ATFP, the

measured reaction eﬃciency is 0.53 0.25 (mean standard

deviation). The high standard deviation illustrates that this data

is much less reproducible than all of the other reaction

eﬃciencies that were measured. For this reaction only, the

eﬃciency varied with pressure and was a relatively poor ﬁt to

the sum of two exponentials. This data suggested that the ion

population was changing in composition as we were studying it.

Thus, data for the reactions of ATFP with deprotonated

tyrosine ions produced from both protic and aprotic solvents

suggest that a gas-phase conversion between carboxylate and

phenoxide ion structures can occur.

Deprotonated Tyrosine Structure as a Function of

Hexapole Accumulation Time. The ratio of carboxylate to

phenoxide ions was studied as a function of accumulation time

for ions in the hexapole that immediately follows our ESI

source. Previously, Kass and co-workers⁹had attributed the

inconsistency between their TMSN₃and PES experiments to

diﬀerent ESI source conﬁgurations, with shorter accumulation

times (100 ms) favoring a carboxylate structure and longer

times (1−5 s) favoring a phenoxide structure. Our mass

spectrometer has a hexapole for ion accumulation and its time

scale can be easily manipulated, allowing us to study

accumulation times similar to those for the two instruments

used by Kass in the PES and TMSN₃experiments.

Figure 3. Reactant loss curve for the reaction of deprotonated tyrosine

with 4-amino-2,3,5,6-tetraﬂuoropyridine (ATFP), which is present at a

constant pressure of 5.6 × 10⁻⁸mbar. The solvent is 99:1 (v/v)

CH₃CN:H₂O. The logarithm of [Tyr−H]⁻intensity is plotted as a

function of reaction time. The experimental data points (black circles)

are ﬁtted to an equation involving the sum of two exponential decays.

carboxylate structure. Therefore, this is the lowest energy

deprotonated tyrosine with a carboxylate structure that our

calculations had previously predicted. The G3(MP2) calculated

GA involving the folded carboxylate, structure A of Figure 1a, is

330.4 kcal/mol. This value is 5.7 kcal/mol higher than the

experimental GA, which is a greater deviation than we typically

ﬁnd between experiment and theory. However, the calculations

indicate that this structure is highly folded with the

deprotonated carboxylate group participating in two hydrogen

bonds. Due to its highly folded structure, the experimental GA

value relating to this structure is likely to have more error than

the other experimental GA values obtained this study because,

as noted above, steric hindrance of the reactive site can yield

erroneously low reaction eﬃciencies.^21,64,65Still, reactions with

TMSN₃show that carboxylate ions are produced almost

exclusively from aprotic solvents, while proton transfer

reactions ﬁnd a carboxylate structure to be the lowest energy

gas-phase conformer or isomer of deprotonated tyrosine

The diﬀerent deprotonated tyrosine ion structures generated

from protic versus aprotic solvents may relate to a diﬀerent

tyrosine neutral structure in the two types of solvents. In

aqueous solution, amino acids exist in their zwitterionic form in

∼10 000 times greater abundance than their non-zwitterion

form. In general, the zwitterionic forms of amino acids

dominate in protic solvents, where they are stabilized by

strong hydrogen bonding to the solvent.⁶⁷Therefore, in our

protic solvent system of CH₃OH:H₂O:NH₄OH, neutral

tyrosine is a zwitterion. Consequently, during the ESI process,

the site that loses a proton to form deprotonated tyrosine is the

ammonium group of the zwitterion. In contrast, for aprotic

solvents, there is no consistency with regard to whether a

neutral amino acid exists in zwitterionic or non-zwitterionic

form. The form relates to both the solvent and the amino acid,

and is diﬃcult to predict. In general, the larger the side chain of

an amino acid, the less likely it is to exist as a zwitterion in

aprotic solvents because steric hindrance will limit solva-

tion.^67,68Due to the low solubility of tyrosine in most solvents,

almost nothing is known about the form of its neutral in

nonaqueous solvents. However, tyrosine has one of the largest

amino acid side chains, which suggests that its neutral exists as a

non-zwitterion in aprotic solvents. Therefore, it is likely that, in

an aprotic solvent, the ESI process is deprotonating the most

Our proton transfer reactions that bracketed GAs in both

protic and aprotic solvent systems used a hexapole accumu-

12638

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

lation time of 0.6−0.7 s. This time was selected because it

generally provided the optimal signal intensity for deprotonated

tyrosine. In ESI, ions are produced continually but the FT-ICR

mass analyzer requires a pulsed packet of ions. To convert a

continuous stream of ions into a pulse, ions leaving the ESI

source are trapped in the hexapole where they remain until

moved into the FT-ICR cell. During the entire accumulation

time, the hexapole can accept new ions. Thus, some ions are

trapped in the hexapole for almost the entire accumulation time

while other more recently formed ions only spend a few

milliseconds in the hexapole.

Because proton transfer reactions with ethyl cyanoacetate can

readily distinguish the carboxylate and phenoxide tyrosine

structures generated from protic solvent systems, reactions with

this reference compound were used to monitor the ion

populations as a function of hexapole accumulation time. With

an extremely short accumulation time of 1 ms, the faster

reacting species (i.e., carboxylate) nearly disappears, resulting in

7% carboxylate ions and 93% phenoxide ions. With an

accumulation time of 0.7 s (from the data of Table 1), there

is 24% carboxylate and 76% phenoxide. For a very long

accumulation time of 5 s, the ratios of the two populations

remained at ∼24:76.

The eﬀects of accumulation time for ions produced from

various solvents were also studied using TMSN₃reactions to

measure the carboxylate and phenoxide abundances. Data

obtained for three solvent systems is shown in Figure 4. Figure

4a involves a protic solvent system of 74.5:24.5:1 (v/v/v)

CH₃OH:H₂O:NH₄OH. At very short accumulation times, the

ion population is 55% phenoxide and 45% carboxylate, but this

stabilizes at ∼70% phenoxide and ∼30% carboxylate after ∼0.7

s. Similar eﬀects were observed with other protic solvent

systems. Figure 4c shows data obtained with an aprotic solvent

of 99:1 (v/v) CH₃OH:H₂O. At short accumulation times, the

ions are almost exclusively carboxylate; at a long time of 5 s,

∼3% of the ions are phenoxide. Figure 4b, involving 98:2 (v/v)

CH₃OH:H₂O, is especially interesting because within less than

1 s of accumulation time the ion ratio goes from ∼55%

carboxylate/45% phenoxide to the reverse. This experiment

was repeated several times over a period of months, and the

result was always the same.

Figure 4. Percentage of product ions from the reaction of

deprotonated tyrosine with trimethylsilyazide (TMSN₃) as a function

of time that the ions accumulate in the hexapole. The presence of m/z

252 indicates a phenoxide deprotonated ion structure, while m/z 42

indicates a carboxylate structure. The solvent systems are (a)

74.5:24.5:1 CH₃OH:H₂O:NH₄OH, (b) 98:2 CH₃CN:H₂O, and (c)

99:1 CH₃CN:H₂O.

energy would be required to convert between the deprotonated

tyrosine isomers and conformers.

The general trend is that, as the ions spend more time in the

hexapole, the relative abundance of phenoxide ions increases

and carboxylate ions decreases. This is especially pronounced

when the ESI solvent system contains at least 2% of a protic

solvent such as H₂O or CH₃OH. A possibility is a preferential

retention of phenoxide ions or a preferential loss of carboxylate

ions, although it is diﬃcult to envision a reason for either of

these events to occur. Another possibility is that a conversion

between structures is occurring. Presumably, this would be the

less stable carboxylate ion (experimental GA of 333.5 kcal/mol)

converting to the slightly more stable phenoxide ion

(experimental GA of 332.4 kcal/mol), just as these ions appear

to convert during the reactions with the reference compound

ATFP.

Hexapole ion guides are generally known for collisional

cooling of ions, but a slight heating of accumulated ions can

occur due to a radial stratiﬁcation eﬀect. This eﬀect is most

pronounced for ions of high charge and high mass-to-charge

ratio.⁷⁰⁻⁷²Deprotonated tyrosine ions are singly charged and of

relatively low mass, making addition of energy during the

hexapole trapping process less likely. However, very little

If a conversion is occurring, it may be facilitated by solvent

interactions with ions in the hexapole. The pressure in the

hexapole is in the 10⁻³mbar range, with the primary

component being dried air (the ESI drying and nebulizer

gas) along with an unknown amount of solvent.⁷³The fact that

the change in carboxylate/phenoxide ratio is most pronounced

in the presence of H₂O supports the involvement of protic

solvent (which can hydrogen bond to tyrosine) in an

isomerization process. Tian and Kass⁸proposed that there

was conversion between the two isomeric structures in the

presence of CH₃OH. They attributed this eﬀect to a gas-phase

relay mechanism^74,75in which CH₃OH simultaneously

coordinates to the deprotonated carboxylic acid and neutral

phenol sites, protonating the carboxylate while deprotonating

the phenol.

We studied structures for the starting point of the proposed

relay mechanism at the DFT level (B3LYP/aug-cc-pVDZ) for

the folded and unfolded carboxylate and phenoxide structures

with the addition of 2, 3, or 6 waters. Relative energies were

calculated for ΔG_gasand ΔG_aqand are given in Table 5. (The

optimized structures of the solvated tyrosine anions are shown

12639

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

Table 5. Relative Free Energies in kcal/mol of Solvated Tyrosine

a

free energy type

no. of waters

folded carboxylate

unfolded carboxylate

unfolded phenoxide

folded phenoxide

ΔG_gas

0

2

3

6

1.0

1.5

1.6

1.4

0.0

2.5

0.0

2.5

0.0

4.0

0.0

6.5

0.0

6.2

0.0

0.5

0.0

1.3

6.4

6.5

4.3

5.8

4.0

8.8

4.4

8.4

10.9

11.0

10.9

6.0

ΔΔG_solv(H₂O)

ΔΔG_solv(CH₃OH)

ΔΔG_solv(CH₃CN)

ΔΔG_gas

ΔΔG_solv(H₂O)

ΔΔG_gas

ΔΔG_solv(H₂O)

ΔΔG_gas

ΔΔG_solv(H₂O)

8.5

7.1

8.7

7.8

8.1

a

ΔΔG_gas= relative free energy in the gas phase. ΔΔG_solv(H₂O) = relative free energy in aqueous solution modeled by a self-consistent reaction ﬁeld.

ΔΔG_solv(CH₃OH) = relative free energy in methanol solution modeled by a self-consistent reaction ﬁeld. ΔΔG_solv(CH₃CN) = relative free energy in

acetonitrile solution modeled by a self-consistent reaction ﬁeld.

in the Supporting Information.) The Gibbs free energy for

deprotonation in aqueous solution (ΔG_aq) was calculated from

the gas-phase free energy and the aqueous solvation free energy

obtained from self-consistent reaction ﬁeld calculations.⁷⁶The

solvation energy was calculated as the sum of the electrostatic

energies (polarized solute−solvent) and the nonelectrostatic

energies using the COSMO parametrization⁷⁷at the B3LYP/

aug-cc-pVDZ level using the gas-phase geometries obtained at

this level. Solvents used include H₂O, CH₃OH, and CH₃CN

with respective dielectric constants of 78.39, 32.63, and 36.64.⁷⁸

In the gas phase, the folded carboxylate and unfolded

phenoxide anions are ∼1 kcal/mol higher in energy than the

unfolded carboxylate structure. Without the addition of any

explicit waters of solvation, the phenoxide anions are still higher

in energy than the carboxylate anions independent of the

solvent used. When using a two- or three-water relay

mechanism, the folded carboxylate anion is more stable in

the gas phase, whereas the unfolded carboxylate is more stable

in aqueous solution. In both cases, the phenoxide anions are

higher in energy than the carboxylate anions by ∼4−9 kcal/

mol. Extra water molecules were added to better solvate the O⁻

and COO⁻sites on the anions. The addition of six waters

resulted in the unfolded and folded carboxylate structures

having approximately the same energy in the gas phase, whereas

in aqueous solution the unfolded carboxylate was still more

stable. The unfolded phenoxide structure is more stable with

the additional waters and has approximately the same relative

energy as the folded carboxylate anion, ∼4 kcal/mol. The

stability in solution of the unfolded structure is likely due to an

eﬀective larger dipole moment than that for the folded

structure. These results show that there is additional chemistry

occurring, leading to the experimental observations.

second lowest energy structure found experimentally, which is

several kcal/mol higher in energy, is deprotonated at the

phenol group and is present when the solvent system contains

at least 2% of a protic compound. In protic solvent, neutral

tyrosine exists as a zwitterion. The third structure is a

carboxylate ion, which G3(MP2) calculations indicate to be

unfolded. This carboxylate ion is only ∼1.5 kcal/mol higher in

energy than the phenolate ion, and the two structures appear to

be converting during ion/molecule reactions with some acidic

reference compounds and also during ion accumulation in a

hexapole. G3(MP2) calculations found that the three structures

diﬀer in energy by ∼2.5 kcal/mol, yet all three are readily

distinguished experimentally by proton transfer ion/molecule

reactions. Experimental conditions, such as solvent and time

that ions are accumulated in a hexapole, greatly aﬀect the

abundance ratios of the deprotonated tyrosine structures. In

contrast, the aromatic amino acid phenylalanine, which has a

benzyl side chain, was found experimentally to have one

structure when electrosprayed from a protic solvent. This is

consistent with the results of G3(MP2) calculations, which

indicate that a folded carboxylate structure is only 0.3 kcal/mol

more stable than an unfolded carboxylate structure.

ASSOCIATED CONTENT

■

S

* Supporting Information

Complete references for refs 35, 48, and 49. Additional

information from the electronic structure calculations including

Cartesian x, y, and z coordinates in Å of the lowest energy

G3MP2MP2(full)/6-31G(d) and B3LYP/aug-cc-pvdz opti-

mized geometries, total H₂₉₈and G₂₉₈energies from Gaussian

09 for all neutrals and anions at the G3(MP2) and B3LYP/aug-

cc-pvdz levels, CCSD(T) atomization energies used for the

benchmark study of the acidity of formic acid, energy

diﬀerences of o-, m-, and p-cresol anions at diﬀerent Gx

computational levels, the multiple low energy conformers of the

neutral amino acids, solvated tyrosine anions, and higher energy

deprotonation sites for the amino acids. This material is

available free of charge via the Internet at http://pubs.acs.org.

CONCLUSIONS

■

Gas-phase deprotonation of the amino acid tyrosine produces

three structures: a species deprotonated at the phenolic side

chain and two conformers deprotonated at the C-terminal

carboxylic acid group. To the best of our knowledge, this is the

ﬁrst time that three distinct structures have been observed

when removing a proton from the same small organic precursor

molecule by mass spectrometry. The lowest energy, most stable

structure is deprotonated at the carboxylic acid group, and

G3(MP2) calculations indicate that the structure is highly

folded with extensive hydrogen bonding. This structure was

only observed experimentally during ESI from aprotic solvents,

where tyrosine may exist in a non-zwitterionic form. The

AUTHOR INFORMATION

■

Present Address

^†(S.S.B.) National Jewish Health, 1400 Jackson Street,

Goodman K923, Denver, CO 80206-2761.

Notes

The authors declare no competing ﬁnancial interest.

12640

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

(13) O’Hair, R. A. J.; Bowie, J. H.; Gronert, S. Gas-Phase Acidities of

the Alpha-Amino Acids. Int. J. Mass Spectrom. Ion Processes 1992, 117,

23−36.

(14) Jones, C. M.; Bernier, M.; Carson, E.; Colyer, K. E.; Metz, R.;

Pawlow, A.; Wischow, E. D.; Webb, I.; Andriole, E. J.; Poutsma, J. C.

Gas-Phase Acidities of the 20 Protein Amino Acids. Int. J. Mass

Spectrom. 2007, 267, 54−62.

(15) Bouchoux, G.; Salpin, J. Y. Re-Evaluated Gas Phase Basicity and

Proton Affinity Data from the Thermokinetic Method. Rapid Commun.

Mass Spectrom. 1999, 13, 932−936.

(16) de Koning, L. J.; Nibbering, N. M. M.; van Orden, S. L.;

Laukien, F. H. Mass Selection of Ions in a Fourier Transform Ion

Cyclotron Resonance Trap using Correlated Harmonic Excitation

Fields (CHEF). Int. J. Mass Spectrom. Ion Processes 1997, 165/166,

209−219.

(17) Bartmess, J. E. In Negative Ion Energetics Data; Linstrom, P. J.,

Ed.; NIST Chemistry WebBook, NIST Standard Reference Database

Number 69; National Institute of Standards and Technology:

Gaithersburg, MD; http://webbook.nist.gov, (retrieved June 11,

2014).

(18) Zhang, K.; Cassady, C. J.; Chung-Phillips, A. Ab Initio Studies of

Neutral and Protonated Triglycines: Comparison of Calculated and

Experimental Gas-Phase Basicity. J. Am. Chem. Soc. 1994, 116, 11512−

11521.

(19) Bartmess, J. E.; Georgiadis, R. M. Empirical Methods for

Determination of Ionization Gauge Relative Sensitivities for Different

Gases. Vacuum 1983, 33, 149−153.

(20) Miller, K. J. Additivity Methods in Molecular Polarizability. J.

Am. Chem. Soc. 1990, 112, 8533−8542.

(21) Zhang, K.; Zimmerman, D. M.; Chung-Phillips, A.; Cassady, C.

J. Experimental and Ab Initio Studies of the Gas-Phase Basicities of

Polyglycines. J. Am. Chem. Soc. 1993, 115, 10812−10822.

(22) Gong, S.; Camara, E.; He, F.; Green, M. K.; Lebrilla, C. B. Chiral

Recognition and the Deprotonation Reaction of Gas-Phase

Cytochrome c Ions. Int. J. Mass Spectrom. 1999, 185−187, 401−412.

(23) Jankiewicz, B. J.; Vinueza, N. R.; Kirkpatrick, L. M.; Gallardo, V.

ACKNOWLEDGMENTS

■

This research is supported by NSF under CHE-0848470, which

is funded by the American Recovery and Reinvestment Act

(ARRA) and is jointly sponsored by NSF’s Analytical & Surface

and Experimental Physical Chemistry sections, and CHE-

1308348. S.S.B. thanks the University of Alabama’s National

Alumni Association for a License Tag Fellowship. M.L.S. and

C.E.P. thank the U.S. Department of Education for a Graduate

Assistance in Areas of National Need fellowship (DoEd-

GAANN grant P200A100190). M.L.S. also thanks the Howard

Hughes Medical Institute (HHMI) for a teaching fellowship

through the Precollege and Undergraduate Science Education

Program. D.A.D. thanks the Robert Ramsay Fund from the

University of Alabama for partial support. The assistance of Dr.

Ken Belmore and Dr. Russell Timkovich with the NMR

experiments is gratefully acknowledged. Some of the CCSD(T)

calculations were performed at the Molecular Sciences

Computing Facility in the W. R. Wiley Environmental

Molecular Sciences Laboratory, a national scientiﬁc user facility

sponsored by DOE’s Oﬃce of Biological and Environmental

Research and located at Paciﬁc Northwest National Laboratory,

operated for the DOE by Battelle.

REFERENCES

■

(1) Li, Z.; Matus, M. H.; Velazquez, H. A.; Dixon, D. A.; Cassady, C.

J. Gas-Phase Acidities of Aspartic Acid, Glutamic Acid, and their

Amino Acid Amides. Int. J. Mass Spectrom. 2007, 265, 213−223.

(2) Tian, Z.; Pawlow, A.; Poutsma, J. C.; Kass, S. R. Are Carboxyl

Groups the Most Acidic Sites in Amino Acids? Gas-Phase Acidity, H/

D Exchange Experiments, and Computations on Cysteine and its

Conjugate Base. J. Am. Chem. Soc. 2007, 129, 5403−5407.

(3) Stover, M. L.; Jackson, V.; Matus, M.; Adams, M.; Cassady, C. J.;

Dixon, D. A. Fundamental Thermochemical Properties of Amino

Acids: Gas-Phase and Aqueous Acidities and Gas-Phase Heats of

Formation. J. Phys. Chem. B 2012, 116, 2905−2916.

(4) DeBlase, A. F.; Kass, S. R.; Johnson, M. A. On the Character of

the Cyclic Ionic H-Bond in Cryogenically Cooled Deprotonated

Cysteine. Phys. Chem. Chem. Phys. 2014, 16, 4569−4575.

(5) Oomens, J.; Steill, J. D.; Redlich, B.; Gas-Phase, I. R.

Spectroscopy of Deprotonated Amino Acids. J. Am. Chem. Soc.

2009, 131, 4310−4319.

A.; Li, G.; Nash, J. J.; Kenttamaa, H. I. Does the 2,6-

̈

Didehydropyridinium Cation Exist? J. Phys. Org. Chem. 2013, 26,

707−714.

(24) He, F.; Marshall, A. G.; Freitas, M. A. Assignment of Gas-Phase

Dipeptide Amide Hydrogen Exchange Rate Constants by Site-Specific

Substitution: GlyGly. J. Phys. Chem. B 2001, 105, 2244−2249.

(25) Watkins, M. A.; Nelson, E. D.; Tichy, S. E.; Kenttamaa, H. I.

̈

Rearrangement of Phenylcarbene Radical Cation to Dehydrotropylium

Cation. Int. J. Mass Spectrom. 2006, 249−250, 1−7.

(6) Bokatzian-Johnson, S. S.; Stover, M. L.; Dixon, D. A.; Cassady, C.

J. Gas-Phase Deprotonation of the Peptide Backbone for Tripeptides

and their Methyl Esters with Hydrogen and Methyl Side Chains. J.

Phys. Chem. B 2012, 116, 14844−14858.

(26) Schaaff, T. G.; Stephenson, J. L.; McLuckey, S. A. Gas Phase H/

D Exchange Kinetics: DI Versus D2O. J. Am. Soc. Mass Spectrom. 2000,

11, 167−171.

(27) Gross, D. S.; Schnier, P. D.; Rodriguez-Cruz, S. E.; Fagerquist,

C. K.; Williams, E. R. Conformations and Folding of Lysozyme Ions in

Vacuo. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 3143−3148.

(28) Cassady, C. J.; Carr, S. R. Elucidation of Isomeric Structures for

Ubiquitin [M+12H]¹²⁺Ions Produced by Electrospray Ionization

Mass Spectrometry. J. Mass Spectrom. 1996, 31, 247−254.

(29) Ewing, N. P.; Cassady, C. J. Effects of Cysteic Acid Groups on

the Gas-Phase Reactivity and Dissociation of [M+4H]⁴⁺Ions from

Insulin Chain B. J. Am. Soc. Mass Spectrom. 1999, 10, 928−940.

(30) Su, T.; Chesnavich, W. J. Parametrization of the Ion-Polar

Molecule Collision Rate-Constant by Trajectory Calculations. J. Chem.

Phys. 1982, 76, 5183−5185.

(31) Su, T. Erratum: Trajectory Calculations of Ion-Polar Molecule

Capture Rate Constants at Low Temperatures. J. Chem. Phys. 1988,

89, 5355−5355.

(32) Bouchoux, G.; Salpin, J. Y.; Leblanc, D. A Relationship Between

the Kinetics and Thermochemistry of Proton Transfer Reactions in

the Gas Phase. Int. J. Mass Spectrom. Ion Processes 1996, 153, 37−48.

(33) Bouchoux, G.; Salpin, J. Y. Gas-Phase Basicity of Glycine,

Alanine, Proline, Serine, Lysine, Histidine and Some of their Peptides

(7) Gao, J.; Cassady, C. J. Negative Ion Production from Peptides

and Proteins by Matrix-Assisted Laser Desorption/ionization Time-of-

Flight Mass Spectrometry. Rapid Commun. Mass Spectrom. 2008, 22,

4066−4072.

(8) Tian, Z.; Kass, S. R. Does Electrospray Ionization Produce Gas-

Phase Or Liquid-Phase Structures? J. Am. Chem. Soc. 2008, 130,

10842−10843.

(9) Tian, Z.; Wang, X. B.; Wang, L. S.; Kass, S. R. Are Carboxyl

Groups the most Acidic Sites in Amino Acids? Gas-Phase Acidities,

Photoelectron Spectra, and Computations on Tyrosine, p-Hydrox-

ybenzoic Acid, and their Conjugate Bases. J. Am. Chem. Soc. 2009, 131,

1174−1181.

(10) Bartlett, R. J.; Musial, M. Coupled-Cluster Theory in Quantum

Chemistry. Rev. Mod. Phys. 2007, 79, 291−352.

(11) Kendall, R. A.; Dunning, T. H., Jr.; Harrison, R. J. Electron

Affinities of the First-Row Atoms Revisited. Systematic Basis Sets and

Wave Functions. J. Chem. Phys. 1992, 96, 6796−6806.

(12) Li, H.; Lin, Z.; Luo, Y. Gas-Phase IR Spectroscopy of

Deprotonated Amino Acids: Global or Local Minima? Chem. Phys.

Lett. 2014, 598, 86−90.

12641

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

by the Thermokinetic Method. Eur. J. Mass Spectrom. 2003, 9, 391−

402.

(53) Curtiss, L. A.; Redfern, P. C.; Raghavachari, K. Gaussian-4

Theory using Reduced Order Perturbation Theory. J. Chem. Phys.

2007, 127, 124105/1−124105/8.

(54) Cumming, J. B.; Kebarle, P. Summary of Gas Phas Acidity

Measurements Involving Acids AH. Entropy Changes in Proton

Transfer Reactions Involving Negative Ions. Bond Dissociation

Energies D(A-H) and Electron Affiniites EA(A). Can. J. Chem.

1978, 56, 1−9.

(34) Bouchoux, G.; Buisson, D.; Bourcier, S.; Sablier, M. Application

of the Kinetic Method to Bifunctional Bases ESI Tandem Quadrupole

Experiments. Int. J. Mass Spectrom. 2003, 228, 1035−1054.

(35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;

Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,

B.; Petersson, G. A.; et al. Gaussian 09, revision B.0.; Gaussian, Inc.:

Wallington, CT, 2010.

(36) Becke, A. D. Density-Functional Thermochemistry. III. The

Role of Exact Exchange. J. Chem. Phys. 1993, 98, 5648−5652.

(37) Parr, R. G.; Yang, W. In Density Functional Theory of Atoms and

Molecules; Oxford University Press: New York, 1989.

(38) Godbout, N.; Salahub, D. R.; Andzelm, J.; Wimmer, E.

Optimization of Gaussian-Type Basis Sets for Local Spin Density

Functional Calculations. Part I. Boron through Neon, Optimization

Technique and Validation. Can. J. Chem. 1992, 70, 560−571.

(39) Gutowski, K. E.; Dixon, D. A. Ab Initio Prediction of the Gas-

and Solution-Phase Acidities of Strong Bronsted Acids: The

Calculation of pKa Values Less than −10. J. Phys. Chem. A 2006,

110, 12044−12054.

(55) Angel, L. A.; Ervin, K. M.; Gas-Phase Acidities; Bond, O.-H.

Dissociation Enthalpies of Phenol, 3-Methylphenol, 2,4,6-Trimethyl-

phenol, and Ethanoic Acid. J. Phys. Chem. A 2006, 110, 10392−10403.

(56) Harrison, A. G.; Tu, Y. Site of Protonation of N-Alkylanilines.

Int. J. Mass Spectrom. 2000, 195/196, 33−43.

(57) Harrison, A. G. Chemical Ionization Mass Spectrometry; CRC

Press: Boca Raton, FL, 1992.

(58) Lau, Y. K.; Kebarle, P. Substituent Effects on the Intrinsic

Basicity of Benzene: Proton Affinities of Substituted Benzenes. J. Am.

Chem. Soc. 1976, 98, 7452−7453.

(59) McLuckey, S. A.; Cameron, D.; Cooks, R. G. Proton Affinities

from Dissociations of Proton-Bound Dimers. J. Am. Chem. Soc. 1981,

103, 1313−1317.

(60) Cheng, X.; Wu, Z.; Fenselau, C. Collision Energy Dependence

of Proton-Bound Dimer Dissociation: Entropy Effects, Proton

Affinities, and Intramolecular Hydrogen Bonding in Protonated

Peptides. J. Am. Chem. Soc. 1993, 115, 4844−4848.

(61) Haupert, L. J.; Poutsma, J. C.; Wenthold, P. G. The Curtin-

Hammett Principle in Mass Spectrometry. Acc. Chem. Res. 2009, 42,

1480−1488.

(62) Fournier, F.; Afonso, C.; Fagin, A. E.; Gronert, S.; Tabet, J. C.

Can Cluster Structure Affect Kinetic Method Measurements? the

Curious Case of Glutamic Acid’s Gas-Phase Acidity. J. Am. Soc. Mass

Spectrom. 2008, 19, 1887−1896.

(63) Jia, B.; Angel, L. A.; Ervin, K. M. Threshold Collision-Induced

Dissociation of Hydrogen-Bonded Dimers of Carboxylic Acids. J. Phys.

Chem. A 2008, 112, 1773−1782.

(40) Curtiss, L. A.; Redfern, P. C.; Raghavachari, K.; Rassolov, V.;

Pople, J. A. Gaussian-3 Theory using Reduced Moller-Plesset Order. J.

Chem. Phys. 1999, 110, 4703−4709.

(41) Dixon, D. A.; Feller, D.; Peterson, K. A. In A Practical Guide to

Reliable First Principles Computational Thermochemistry Predictions

Across the Periodic Table; Wheeler, R. A., Section Ed., Tschumper, G.

S., Eds.; Annual Reports in Computational Chemistry; Elsevier:

Amsterdam, The Netherlands, 2012; Vol. 8, pp 1−28.

(42) Feller, D.; Peterson, K. A.; Dixon, D. A. Further Benchmarks of

a Composite, Convergent, Statistically Calibrated Coupled-Cluster-

Based Approach for Thermochemical and Spectroscopic Studies. Mol.

Phys. 2012, 110, 2381−2399.

(43) Peterson, K. A.; Feller, D.; Dixon, D. A. Chemical Accuracy in

Ab Initio Thermochemistry and Spectroscopy: Current Strategies and

Future Challenges. Theor. Chem. Acc. 2012, 131, 1−20.

(44) Feller, D.; Peterson, K. A.; Dixon, D. A. A Survey of Factors

Contributing to Accurate Theoretical Predictions of Atomization

Energies and Molecular Structures. J. Chem. Phys. 2008, 129, 204105/

1−204105/32.

(64) Buker, H.-H.; Grutzmacher, H.-F. A Fourier Transform-Ion

Cyclotron Resonance Study of Steric Effects on Proton Transfer

Reactions of Polyalkyl Benzenes. Int. J. Mass Spectrom. Ion Processes

1991, 109, 95−104.

(65) Attina, M.; Cacace, F. Concerning Sterical Hindrance to

Deprotonation of Gaseous Polyalkylbenzenium Ions. Int. J. Mass

Spectrom. Ion Processes 1992, 120, R1−R4.

(45) Feller, D.; Dixon, D. A. Extended Benchmark Studies of

Coupled Cluster Theory through Triple Excitations. J. Chem. Phys.

2001, 115, 3484−3496.

(46) McQuarrie, D. A. Statistical Mechanics; Harper & Row: New

York, 1976.

(66) Schroder, D.; Budesinsky, M.; Roithova, J. Deprotonation of p-

̈

Hydroxybenzoic Acid: Does Electrospray Ionization Sample Solution

or Gas-Phase Structures? J. Am. Chem. Soc. 2012, 134, 15897−15905.

(67) Hughes, D. L.; Bergan, J. J.; Grabowski, E. J. J. Amino Acid

Chemistry in Dipolar Aprotic Solvents: Dissociation Constants and

Ambident Reactivity. J. Org. Chem. 1986, 51, 2579−2585.

(68) Cheung, E. T. Substituent Eﬀects on the Tautomerization of

Amino Acids. MS Thesis, Texas Tech University, Lubbock, TX, 1995.

(69) Lide, D. R., Ed. CRC Handbook of Chemistry and Physics; CRC

Press: Cleveland, OH, 1992−1993.

(47) Muftakhov, M. V.; Vasil’ev, Y. V.; Mazunov, V. A.

Determination of Electron Affinity of Carbonyl Radicals by Means

of Negative Ion Mass Spectrometry. Rapid Commun. Mass Spectrom.

1999, 13, 1104−1108.

(48) Knowles, P. J.; Manby, F. R.; Schutz, M.; Celani, P.; Knizia, G.;

̈

Korona, T.; Lindh, R.; Mitrushenkov, A.; Rauhut, G.; Adler, T. B.;

et al. MOLPRO, version 2010.1, A Package of Ab Initio Programs;

University College Cardiﬀ Consultants Limited: Cardiﬀ, Wales, UK,

2010; http://www.molpro.net/.

(70) Tolmachev, A. V.; Udseth, H. R.; Smith, R. D. Radial

Stratification of Ions as a Function of Mass to Charge Ratio in

Collisional Cooling Radio Frequency Multipoles used as Ion Guides or

Ion Traps. Rapid Commun. Mass Spectrom. 2000, 14, 1907−1913.

(71) McDonnell, L. A.; Giannakopulos, A. E.; Derrick, P. J.; Tsybin,

Y. O.; Hakansson, P. A. Theoretical Investigation of the Kinetic Energy

of Ions Trapped in a Radio-Frequency Hexapole Ion Trap. Eur. J. Mass

Spectrom. 2002, 8, 181−189.

(72) Tolmachev, A. V.; Udseth, H. R.; Smith, R. D. Modeling the Ion

Density Distribution in Collisional Cooling RF Multipole Ion Guides.

Int. J. Mass Spectrom. 2003, 222, 155−174.

(73) Easterling, M. L., Bruker Daltonics FT-ICR Product Manager.

Private correspondence, March 28, 2012.

(49) Valiev, M.; Bylaska, E. J.; Govind, N.; Kowalski, K.; Straatsma,

T. P.; Van Dam, H. J. J.; Wang, D.; Nieplocha, J.; Apra, E.; Windus, T.

L.; et al. NWChem: A Comprehensive and Scalable Open-Source

Solution for Large Scale Molecular Simulations. Comput. Phys.

Commun. 2010, 181, 1477−1489.

(50) Curtiss, L. A.; Raghavachari, K. Gaussian-3 (G3) Theory for

Molecules Containing First and Second-Row Atoms. J. Chem. Phys.

1998, 109, 7764−7776.

(51) Baboul, A. G.; Curtiss, L. A.; Redfern, P. C.; Raghavachari, K.

Gaussian-3 Theory using Density Functional Geometries and Zero-

Point Energies. J. Chem. Phys. 1999, 110, 7650−7657.

(52) Curtiss, L. A.; Redfern, P. C.; Raghavachari, K. Gaussian-4

Theory. J. Chem. Phys. 2007, 126, 084108/1−084108/12.

(74) Campbell, S.; Rodgers, M.; Marzluff, E.; Beauchamp, J.

Deuterium Exchange Reactions as a Probe of Biomolecule Structure.

Fundamental Studies of Gas Phase H/D Exchange Reactions of

12642

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

The Journal of Physical Chemistry B

Article

Protonated Glycine Oligomers with D₂O, CD₃OD, CD₃CO₂D, and

ND₃. J. Am. Chem. Soc. 1995, 117, 12840−12854.

(75) Gard, E.; Green, M. K.; Bregar, J.; Lebrilla, C. B. Gas-Phase

Hydrogen/Deuterium Exchange as a Molecular Probe for the

Interaction of Methanol and Protonated Peptides. J. Am. Soc. Mass

Spectrom. 1994, 5, 623−631.

(76) Tomasi, J.; Mennucci, B.; Cammi, R. Quantum Mechanical

Continuum Solvation Models. Chem. Rev. 2005, 105, 2999−3094.

(77) Klamt, A.; Schuurmann, G. COSMO: A New Approach to

̈

Dielectric Screening in Solvents with Explicit Expressions for the

Screening Energy and its Gradient. J. Chem. Soc., Perkin Trans. 2 1993,

799−805.

(78) Cammi, R.; Mennucci, B.; Tomasi, J. In Computational Modelling

of the Solvent Eﬀects on Molecular Properties: An Overview of the

Polarizable Continuum Model (PCM) Approach; Leszczynski, J., Ed.;

Computational Chemistry: Reviews of Current Trends; World

Scientiﬁc: Singapore, 2003; Vol. 8, pp 1−79.

(79) Fujio, M.; McIver, R. T.; Taft, R. W. Effects of the Acidities of

Phenols from Specific Substituent-Solvent Interactions. Inherent

Substituent Parameters from Gas-Phase Acidities. J. Am. Chem. Soc.

1981, 103, 4017−4029.

(80) Caldwell, G.; Renneboog, R.; Kebarle, P. Gas-Phase Acidities of

Aliphatic Carboxylic Acids, Based on Measurements of Proton

Transfer Equilibria. Can. J. Chem. 1989, 67, 611−618.

(81) Kim, E. H.; Bradforth, S. E.; Arnold, D. W.; Metz, R. B.;

Neumark, D. M. Study of HCO₂and DCO₂by Negative Ion

Photoelectron Spectroscopy. J. Chem. Phys. 1995, 103, 7801−7814.

(82) Taft, R. W.; Topsom, R. D. The Nature and Analysis of

Substitutent Electronic Effects. Prog. Phys. Org. Chem. 1987, 16, 1−84.

(83) Jinfeng, C.; Topsom, R. D.; Headley, A. D.; Koppel, I.; Mishima,

M.; Taft, R. W.; Veji, S. Acidities of Substituted Acetic Acids. J. Mol.

Struct.: THEOCHEM 1988, 168, 141−146.

(84) Norrman, K.; McMahon, T. B. Intramolecular Solvation of

Carboxylate Anions in the Gas Phase. J. Phys. Chem. A 1999, 103,

7008−7016.

(85) Richardson, J. H.; Stephenson, L. M.; Brauman, J. I.

Photodetachment of Electrons from Phenoxides and Thiophenoxide.

J. Am. Chem. Soc. 1975, 97, 2967−2970.

(86) Bartmess, J. E.; Scott, J. A.; McIver, R. T. Scale of Acidities in

the Gas-Phase from Methanol to Phenol. J. Am. Chem. Soc. 1979, 101,

6046−6056.

(87) Gunion, R. F.; Gilles, M. K.; Polak, M. L.; Lineberger, W. C.

Ultraviolet Photoelectron Spectroscopy of the Phenide, Benzyl and

Phenoxide Anions, with Ab Initio Calculations. Int. J. Mass Spectrom.

Ion Processes 1992, 117, 601−620.

(88) DeTuri, V. F.; Ervin, K. M. Proton Transfer between Cl⁻and

C₆H₅OH. O-H Bond Energy of Phenol. Int. J. Mass Spectrom. Ion

Processes 1998, 175, 123−132.

(89) Angel, L. A.; Ervin, K. M. Competitive Threshold Collision-

Induced Dissociation: Gas-Phase Acidity and O-H Bond Dissociation

Enthalpy of Phenol. J. Phys. Chem. A 2004, 108, 8346−8352.

(90) McMahon, T. B.; Kebarle, P. Intrinsic Acidities of Substituted

Phenols and Benzoic Acids Determined by Gas-Phase Proton-Transfer

Equilibria. J. Am. Chem. Soc. 1977, 99, 2222−2230.

12643

dx.doi.org/10.1021/jp510037c | J. Phys. Chem. B 2014, 118, 12630−12643

Article Doi

An experimental and computational investigation into the gas-phase acidities of tyrosine and phenylalanine: Three structures for deprotonated tyrosine

DOI: 10.1021/jp510037c

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jp510037c

Full text of DOI:10.1021/jp510037c

Products guided by the article

R&D Labs maybe for 5269-46-5

Relevant to this article

Hot Product