310APPLICATIONS TO SMALL MOLECULES
TABLE 7-5. Isotopic Abundances for Ions Containing Different Numbers of Sulfur, Chlorine, and Bromine Atoms
Number Number Number Number Number Number Number
Number of Br of Br of Br of Br of Br of Sulfur of Sulfur
of Cl Atoms Atoms Atoms Atoms Atoms Atoms Atoms
Atoms Mass (0) (1) (2) (3) (4) (1) (2)
0 A 100 51.5 34.3 17.6
A+2 97.3 100 100 68.5
A+4 48.7 97.3 100
A+6 31.6 64.9
A+8 15.8
1 A 100 77.3 44.2 26.5 14.4
A+2 32 100 100 85.8 60.8
A+4 24.1 69.3 100 100
A+6 13.4 48.5 79.4
A+8 7.8 30
A+10 4.1
2 A 100 62 38.7 20.8 12.1
A+2 64 100 100 74 54.8
A+4 10.2 45 88.8 100 100
A+6 6.2 31.2 63.1 93.3
A+8 3.8 18.3 46.4
A+10 2 11.5
A+12 1.1
A 100 100
A+1 0.8 1.6
A+2 4.4 8.8
A+3 0.07
A+4 0.19
molecular ions at m/z 521 ([M + H
+

. The characteristic isotopic patterns resulting from combi-
nations of the isotope peaks can be used to ascertain elemental composition
of the corresponding ion.
7.6.1.3 Nitrogen Rule. If the nominal molecular weight of an analyte
appears to be an even mass number, the compound contains an even number
of nitrogen atoms (or no nitrogen atoms). On the other hand, if the nominal
molecular weight of an analyte appears to be an odd mass number, the com-
pound contains an odd number of nitrogen atoms. This so-called “Nitrogen
Rule” is very useful for determining the nitrogen content of an unknown com-
pound. In the case of Verapamil (Scheme 1), the molecular weight of the com-
pound is 420 Da, indicating an even number of nitrogen atoms in the molecule.
7.6.1.4 Hydrogen/Deuterium (H/D) Exchange. The exchange of hydrogen
for deuterium in organic molecules has been used in mass spectrometry for
structural studies in both solution phase and gas phase. It also has wide appli-
cations in the structural studies of proteins [24].This method measures the dif-
ference in molecular weight of a compound before and after the deuterium
exchange to determine the exchangeable hydrogens in a molecule for struc-
tural elucidations. For example, one can determine the number of labile hydro-
gen atoms from the mass shift X of [M + H
+
] in H
2
O to [M + D
+
] in D
2
O as X
− 1. The exchangeable hydrogens are usually bound to N, O, or S atoms in func-
tional groups such as OH, NH, NH
2

als.The development of modern MS instrumentation (i.e., QTOF, FT-ICR, etc.)
has allowed accurate mass determinations of small molecules as well as bio-
molecules that are present at very low levels.
An internal mass calibration is generally needed to achieve mass measure-
ment accuracy of 5 to 10ppm with a Q-TOF MS analysis [62–64]. Internal cal-
ibration is based on mixing one or several internal standards or calibrants of
known molecular weight with the analyte and then using the known masses
to calibrate the mass measurements of unknowns that coexist in the sample
mixture.
Currently, FT-ICR MS provides the highest mass resolving power and mass
accuracy among all the mass spectrometric methods. Using external calibra-
tion, FT-ICR MS is capable of achieving mass measurement accuracies of 1
ppm or better. Internal calibration can provide an order-of-magnitude greater
mass accuracy than external calibration for the FT-MS analysis.
One example is the high-resolution ESI-FT-ICR MS analysis of a
mixture of Verapamil and peptide A. The following MS data were obtained
for the mono-isotopic molecular ion and the isotopic molecular ions of
peptide A using an external calibration (calculated mass; mass error in
ppm): (829.5393, 0.1 ppm), (830.5423, 1.3 ppm), (831.5450, 1.0 ppm), and
(832.5476, 1.2 ppm). Similar results were obtained for the Verapamil (data not
shown).
Accurate mass measurement is important in establishing compound iden-
tity. For example, an unknown with a mass of m/z 122.0606 ([M + H
+
]) can be
C
7
H
8
NO; it cannot be C

C
x
H
y
N
z
O
n
, is a reasonable elemental composition for a certain mass, one can
calculate the DBE (numbers of rings and double bonds) of the formula. The
calculation is based on the valences of elements involved, as shown in equa-
tion (7-1). For example, pyridine (C
5
H
5
N) has a calculated DBE of 4 (= 5 −
5/2 + 1/2 + 1), indicating the ring and three double bonds in this molecule. For
benzene (C
6
H
6
), its calculated DBE is 4 (= 6 − 6/2 + 1), suggesting the ring and
three double bonds in the molecule as well.
A more general case, I
y
II
n
III
z
IV

charged parent ions are the ions observed at low-mass range of a MS/MS spec-
trum. The fragmentation chemistry and mechanisms are reasonably under-
stood. The prominent abundant fragment ions are the most stable fragments
that tend to be formed. The fragmentation processes also depend on the sta-
bility of the transition states by which the ions are produced.
Many of the fragment ions observed in the product-ion spectra are formed
by collision-induced heterolytic cleavage. For example, the formation of a
product ion, [M + H
+
− HX], can be explained by a 1,4 hydrogen rearrange-
ment mechanism (Scheme 3). The product ion is formed by the neutral loss of
HX, where X can be a heteroatom or a more electronegative group. A less
common fragmentation mechanism by homolytic cleavage is also observed in
tandem MS experiments. In this case, the driving force for the fragmentation
of an ion is dependent on the stabilities of the resulting ion and the radical
species relative to the energy of the initial ionic species. For instance, the for-
mation of stable product ions, including acylium ion, benzylic ion, and allylic
carbonium ion, are able to promote homolytic cleavage.
DBE + Rdb
()
=− +=27 32 2 1 12
MS INTERPRETATION 313
Scheme 3.
Charge-remote fragmentation is defined as a class of gas-phase decompo-
sitions that occur physically remote from the charge site [66–70]. Although the
mechanism of charge-remote fragmentation is still debatable (Scheme 4) [67],
It is possible to derive structural information from the fragmentation
pattern in a spectrum. The appearance of prominent peaks at certain mass
numbers is empirically correlated with certain structural features. For example,
the mass spectrum of an aromatic compound is usually dominated by a peak

Scheme 4.
it has been proven useful in the structural determination of long-chain or poly-
ring molecules, including fatty acids, phospholipids, glycolipids, triacylglycerols,
steroids, peptides, ceramides, and so on.
7.7 PRACTICAL APPLICATIONS
Mass spectrometry is a powerful and effective technology in drug discovery
and development. This section will concentrate on the practical applications
of LC/MS in problem solving, including high-throughput LC/MS analysis for
combinatorial chemistry, structural characterization of impurities and decom-
position products in bulk drug substances, and identification and quantifica-
tion of drug metabolites.
7.7.1 High-Throughput LC/MS for Combinatorial Chemistry
The application of combinatorial chemistry to the synthesis of potential ther-
apeutic agents has received increasing attention such that combinatorial chem-
istry is now an important tool in modern drug discovery [71]. Automated
approaches capable of screening large libraries of small molecules have
resulted in the successful application of LC/MS in combinatorial chemistry.
Current trends for further integration of LC/MS techniques with new
instrumental development have generated structure-based assays for drug
discovery.
To assess the quality of a combinatorial chemistry library, it is essential to
determine the purity and quantity of the expected products. Commercial soft-
ware, developed by instrument manufacturers, has made possible the unat-
tended and rapid analysis of tens of thousands of individual components of a
specific library. The application of LC/MS in high-throughput screening of
combinatorial libraries has been reviewed by several authors [72–78].
An important application of LC/MS in relation to combinatorial synthesis
is the introduction of open-access LC/MS instrumentation. The dedicated
PRACTICAL APPLICATIONS 315
Scheme 5.

nal calibrations, respectively. In another application, Nawrocki et al. [82]
employed a 4.7-T external-source ESI FT-ICR mass spectrometer to analyze
small-peptide libraries, demonstrating the feasibility of analyzing several com-
binatorial libraries containing 100 to 10,000 small peptides. Furthermore, by
comparing the FTMS data with computer-simulated combinatorial library
mass spectra, the authors were able to monitor the diversity and degeneracy
of the library syntheses.
A well-established method for drug discovery is the utilization of a biolog-
ical assay to screen a large library of small organic molecules for their ability
to bind target biopolymers (i.e., protein) in a specific assay [83, 84]. In general,
the MS-based technologies have the advantage that only small amounts of
protein reagent are required. Recently, Annis et al. [85, 86] reported a high-
throughput affinity selection–mass spectrometry assay to screen mass-encoded
2500-member combinatorial libraries. A schematic representation of the
method is shown in Figure 7-12. Combination of the protein and a small
316 APPLICATIONS TO SMALL MOLECULES
molecule library leads to the formation of a complex of the protein with any
suitable library member. Size-exclusion chromatography (SEC) is then
employed to rapidly separate the protein target, along with any small mole-
cules bound to the target, from any unbound small molecules. The SEC band
containing the complex is immediately transferred to a reversed-phase chro-
matography column (60°C and pH 2). This step serves to denature the target,
thereby dissociating the previously bound small molecules from the complex.
The unbound small molecules are directly introduced into a high-resolution
mass spectrometer for analysis. By using the affinity selection–mass spectro-
metry method, Annis and co-workers discovered a bioactive ligand for the
anti-infective target Escherichia coli dihydrofolate reductase (DHFR) [85, 86].
7.7.2 Characterization of Impurities and Decomposition Products in Bulk
Drug Substances
One of the major applications of LC/MS in pharmaceutical analysis is the iden-

levels.
The capabilities of separating a mixture containing highly varied concen-
trations of analyte and structural characterization of impurities have led to the
increased use of LC/MS. Nicolas and Scholz [87] illustrated the characteriza-
tion of a number of DuP 941 (Scheme 6) impurities by LC/MS(/MS). The five
unknown impurities were labeled A, D, E, F, and G along with the known
impurities B and C in the total ion chromatogram of DuP 941 obtained from
LC/MS analysis (Figure 7-13B). Because the UV-visible absorption and
response factor of related compounds tend to be similar, while their MS ion-
ization efficiencies can be significantly different, it is always useful to record
the UV chromatogram (Figure 7-13A) as well as the mass spectra for the iden-
tification of impurities. The protonated molecular ions ([M + H
+
]) of the impu-
rities A, D, E, F, G were found to be at m/z 392, 339, 324, 482, and 558,
respectively. The molecular ions of the impurities were selected for tandem
MS analysis in order to identify the unknown structures (Figure 7-13C). The
product-ion spectra for the five unknown impurities were shown in Figure 7-
14. The impurity A was found to be a by-product (Figure 7-15A) during the
synthetic process. Major fragment ions were observed at m/z 319, 305, and 261
318 APPLICATIONS TO SMALL MOLECULES
Scheme 6. Structure of DuP 941.
(Reprinted from reference 87, with permission of
Elsevier Science.)
(Figure 7-14A). The base peak at m/z 305 might arise from the neutral loss
of a 2-vinylamino-ethanol. The product ions formed by the neutral loss of 73
(2-methyleneamino-ethanol) and 44 (Ethenol) Da from the precursor ion of
m/z 392 were also consistent with the proposed structure (Figure 7-15A). The
production spectra of impurities D and E were similar in two ways. Both have
a base peak at m/z 88 which was produced when an N-(2-hydroxyethyl)

designed to emulate stresses the compound might experience during manu-
facturing processes and storage. These methods exposed drug candidates to
forced degradation conditions such as acid, base, heat, oxidation, and expo-
sure to light. A successful identification of the degradation products can help
formulation scientists to understand the degradation mechanism of drug can-
didate and improve the clinical formulation development.
There had been numerous reports in the literature that involved LC/MS
and LC/MS/MS for characterization of degradation products [1, 88–97]. An
early example of the rapid structure elucidation of drug degradants induced
by acid, base and heat by LC/MS was reported by Rourick et al. in 1996 [90].
In general, the LC/UV/MS provided the UV and molecular weight data,
PRACTICAL APPLICATIONS 321
Figure 7-15. Possible structures of unisolated DuP 941 impurities showing proposed
CID fragmentations. (Reprinted from reference 87, with permission of Elsevier
Science.)
whereas the tandem LC/MS provided substructural information. The same
group demonstrated that the similar procedure could be applied to obtain the
structural information of the degradation products of paclitaxel (Taxol) [89].
Recently, Feng et al. [92] investigated the oxidative degradation products of
an antifungal agent, SCH 56592 (Scheme 7), by both LC/MS and LC/NMR
analysis. Four major oxidative degradation products of SCH 56592 were char-
acterized, and the oxidation was found to be occurred at the piperazine ring
in the center of the drug molecule.
322 APPLICATIONS TO SMALL MOLECULES
Scheme 7. Structure of SCH 56592.
Shipkova et al. [94] demonstrated the use of on-line high-resolution
LC/ESI-MS using a magnetic sector mass spectrometer for analysis of minor
components in complex mixtures. Everninomicin (SCH 27899; Figure 7-16)
belongs to an important group of oligosaccharide antibiotics isolated from the
fermentation broth of Micromonospora carbonaceae. The compound was

carried out LC FTMS and tandem MS experiments for the detection and iden-
tification of various degradants from drug candidates. This approach drasti-
cally reduced the time required for isolation and purification of substantial
quantities of material and expedited the identification process.
PRACTICAL APPLICATIONS 323
Figure 7-17. Reconstructed ion chromatogram (RIC) of a bulk drug substance (SCH
27899) degraded in ammonium hydroxide solution at pH 10, displaying all the identi-
fied mixture components 1–10. (Reprinted from reference 94, with permission of John
Wiley & Sons.)