Electrogenerated chemiluminescence derivatization reagent, 3-isobutyl-9,10-dimethoxy-1,3,4,6,7,11b-hexahydro-2H-pyrido [2,1-a]isoquinolin-2-ylamine, for carboxylic acid in high-performance liquid chromatography using tris(2,2′-bipyridine)ruthenium(II)

Article abstract of DOI:10.1021/ac020377i

3-Isobutyl-9,10-dimethoxy-1,3,4,6,7,11b-hexahydro-2H-pyrido [2,1-a]isoquinolin-2-ylamine (IDHPIA) was found to be a selective and highly sensitive derivatization reagent for carboxylic acid by high-performance liquid chromatography (HPLC) with electrogene

Full text of DOI:10.1021/ac020377i

Anal. Chem. 2003, 75, 940-946
Electrogenerated Chemiluminescence
Derivatization Reagent,
3-Isobutyl-9,10-dimethoxy-1,3,4,6,7,11b-
hexahydro-2H-pyrido[2,1-a]isoquinolin-2-ylamine,
for Carboxylic Acid in High-Performance Liquid
Chromatography Using
Tris(2,2′-bipyridine)ruthenium(II)
Hirotoshi Morita* and Masaharu Konishi
Shionogi Research Laboratories, Shionogi & Co., Ltd., 12-4 Sagisu 5-Chome, Fukushima-ku, Osaka 553-0002, Japan
workers,1-6 and the postulated reaction mechanism is reported
by Zu and Bard to be as follows:4,5
3-Isobutyl-9,10-dimethoxy-1,3,4,6,7,11b-hexahydro-2H-
pyrido[2 ,1 -a ]isoquinolin-2 -ylamine (IDHP IA) was found
to be a selective and highly sensitive derivatization reagent
for carboxylic acid by high-performance liquid chroma-
tography (HP LC) with electrogenerated chemilumines-
cence detection using tris(2 ,2 ′-bipyridine)ruthenium(II).
Free fatty acids and phenylbutylic acid were used as
model compounds of carboxylic acids, and the derivati-
zation conditions were optimized with myristic acid.
Under the mild reaction conditions of room temperature
for 4 5 min in acetonitrile containing 2 -bromo-1 -ethyl-
pyridinium tetrafluoroborate and 9 -methyl-3 ,4 -dihydro-
2H-pyrido[1,2-a ]pyrimidin-2-one, all the fatty acids tested
were reacted with IDHP IA to produce highly sensitive
derivatives. The chemiluminescence intensity was es-
sentially the same for all fatty acids. The derivatives
obtained from 1 0 free fatty acids were completely sepa-
rated by reversed-phase chromatography under isocratic
elution conditions. The on-column detection limit (signal-
to-noise ratio of 3 ) with proposed HP LC separation
and chemiluminescence detection was 0 .5 and 0 .6 fmol
for myristic acid and phenylbutylic acid, respectively.
IDHP IA was 1 0 0 -fold more sensitive than previously
developed reagents (Morita, H.; Konishi, M. An a l. Ch em .
2 0 0 2 , 74, 1 5 8 4 -1 5 8 9 ). The free fatty acids in human
serum were successfully determined using the present
method.
Ru(bpy)32+ f Ru(bpy)33+ + e-
TA f TA•+ + e-
(1)
(2a)
(2b)
Ru(bpy)33+ + TA f Ru(bpy)32+ + TA•+
TA•+ f TA• + H+
(3)
TA• + Ru(bpy)33+ f TA fragment + [Ru(bpy)32+]* (4)
[Ru(bpy)32+]* f Ru(bpy)32+ + hν
(5)
It was thought that Ru(bpy)33+, obtained by electrochemical
oxidation, reacted with tertiary amine (TA) to give cation radical
•+
(TA ). However, cation radical is supposed to be mainly produced
by direct oxidation at the electrode (eq 2a) and catalytic amine
3+
oxidation by Ru(bpy)3 (eq 2b) is less important, recently. It is
thought to produce a neutral radical (TA•) by elimination of an
R-proton. This unstable radical compound immediately reacts with
3+
Ru(bpy)3 to give the excited state of the ruthenium complex,
[Ru(bpy)32+]*. Its emission to the background state gives phos-
phorescence, with the maximal wavelength at 620 nm. On the
other hand, a new and different reaction mechanism has also been
reported by Bard and co-workers quite recently.6 The tertiary
amine group can be selectively detected by this system, and a
variety of drugs and endogenous compounds containing a tertiary
amine have been sensitively determined with simple, minimal
The electrogenerated chemiluminescent (ECL) system using
tris(2,2′-bipyridine)ruthenium(II) [Ru(bpy)32+] has recently be-
come a powerful tool for the determination of compounds that
have tertiary amine or diketone groups because of its high
sensitivity and selectivity. The reaction mechanism between Ru-
(1) Leland, J. K.; Powell, M. J. J. Electrochem. Soc. 1 9 9 0 , 137, 3127-3131.
(2) Liang, P.; Sanchez, R. I.; Martin, M. T. Anal. Chem. 1 9 9 6 , 68, 2426-2431.
(3) Brune, S. N.; Bobbitt, D. R. Anal. Chem. 1 9 9 2 , 64, 166-170.
(4) Zu, Y.; Bard, A. J. Anal. Chem. 2 0 0 0 , 72, 3223-3232.
(5) Zu, Y.; Bard, A. J. Anal. Chem. 2 0 0 1 , 73, 3960-3964.
(6) Miao, W.; Choi, J.-P.; Bard, A. J. J. Am. Chem. Soc. 2 0 0 2 , 124, 14478-
14485.
2+
(bpy)3 and tertiary amine has been investigated by many
* Corresponding author. Phone: +81-6-6458-5861. Fax: +81-6-6458-0987.
E-mail: hirotoshi.morita@shionogi.co.jp.
940 Analytical Chemistry, Vol. 75, No. 4, February 15, 2003
10.1021/ac020377i CCC: $25.00 © 2003 American Chemical Society
Published on Web 01/21/2003

pretreatment of biological samples.7-13 Some applications for high-
performance liquid chromatography (HPLC) or flow injection
analysis (FIA) by Ru(bpy)33+ chemiluminescence (CL) have been
developed in recent years.3,7-23 Nevertheless, the compounds
detectable by this method are still limited. Therefore, we at-
tempted to extend the system to a wider range of compounds
containing nonsuitable functional groups by converting a variety
of compounds into derivatives suitable for ECL detection by
tagging appropriate molecules with an analyte.
Many derivatization reagents for fluorescence, ultraviolet
absorbance, electrochemical, and chemiluminescence detection
have been developed. However, only a few derivatization reagents
for ECL detection such as dansyl chloride21 or divinylsulfone23
have been developed. We have developed some reagents such
as N-(3-aminopropyl)pyrrolidine for carboxylic acid24 and 3-(di-
ethylamino)propionic acid for primary amine24 and alcohol.25 These
reagents reacted with the analytes to produce sensitive derivatives
in the presence of suitable catalysts at room temperature for
30-120 min and showed the detection limit of 40-70 fmol at a
signal-to-noise ratio of 3.
3,4-dihydro-2H-pyrido[1,2-a]pyrimidin-2-one (MDPP) were from
Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Acetonitrile
and methanol of HPLC grade were obtained from Kanto Chemical
Co. Inc. Tetrabenazine was synthesized in this laboratory but is
commercially available. All the other chemicals were of guaranteed
grade and used without further purification. Twice-distilled deion-
ized water was used throughout the study.
Instrumentation. Melting points were recorded on a Yana-
gimoto micro melting point apparatus (Tokyo, Japan) and are
uncorrected. 1H NMR spectra were recorded on an ASM-100 (300
MHz) spectrometer (Varian) using tetramethylsilane as an internal
standard. Mass spectra (electron impact ionization) were deter-
mined with a GCMS-QP1000(A) spectrometer (Shimadzu). High-
resolution (HR) mass spectra (FAB-MS) were determined with a
SX/ SX102A (JEOL) using m-nitrobenzyl alcohol as a matrix.
Elemental analyses were performed by CHN corder MT-6
(Yanako).
The ECL intensity was observed by modifying a commercially
available system. The HPLC system is shown in Figure 1. All
the solutions were purged with two types of degassers (DGU-
10B for helium gas purge type and DGU-3A for membrane type,
Shimadzu) and were delivered with pumps (LC-10AD, Shimadzu).
Electrochemical oxidation was performed with a porous graphite
working electrode (guard cell, model 5020, ESA), and the current
was controlled with a potentio-galvanostat (NPGS-2501, Nikkoh
Keisoku). The chemiluminescence detector used was a Shimadzu
CLD-10A equipped with an 80-µL spiral flow cell and R374HA
photomultiplier tube (Hamamatsu Photonics). Shimadzu SIL-10A
was used as the autosampler. CLASS LC-10 (Shimadzu) was used
as the data processor. To avoid the permeation of atmospheric
oxygen, a metal tube was used for the connection. A low-volume
tee tube was used for mixing of the mobile phase and the reagent
solution.
In this study, we examined a more sensitive derivatization
reagent and developed a 100-fold more sensitive reagent for
carboxylic acid than previously developed reagents.24,25 To our
knowledge, this reagent is one of the most sensitive derivatization
reagents developed for HPLC analysis. Moreover, this reagent
showed the optimum detection pH at 1.5-2.0 in the ECL system,
making it possible to omit the pH-conditioning pump and simplify-
ing the equipment of the detection system.
EXPERIMENTAL SECTION
Reagents. Tris(2,2′-bipyridine)ruthenium(II) chloride penta-
hydrate (Ru(bpy)3Cl2‚5H2O) and ibuprofen were purchased from
Sigma Chemical Co. (St. Louis, MO). Phenylbutylic acid (PB) was
obtained from Kanto Chemical Co. Inc. (Tokyo, Japan). Fatty acids
were obtained from Wako Pure Chemical Indutries, Ltd. (Osaka,
Japan), Sigma Chemical Co., Nacalai Tesque Inc., or GL Sciences
Inc. (Tokyo, Japan). 1,2,3,4,6,7-Hexahydrobenzo[a]quinolizine-2,4-
dione was purchased from Bionet Research Ltd. (Cornwall, U.K.).
2-Bromo-1-ethylpyridinium tetrafluoroborate (BEPT) and 9-methyl-
Synthesis of 1,3,4,6,7,11b-Hexahydro-2H-pyrido[2,1-a ]-
isoquinolin-2 -ol (HP I). To a solution of 1,2,3,4,6,7-hexahydro-
benzo[a]quinolizine-2,4-dione (1 ) (1.86 g, 8.64 mmol) in 25 mL
of ethyl acetate/ methanol (2:3, v/ v) in an ice bath was added
NaBH4 (17.3 mmol), and the mixture was stirred for 2 h at room
temperature. The reaction mixture was evaporated to dryness in
vacuo. The residue was dissolved with chloroform and washed
with water. The organic layer was dried with anhydrous MgSO4,
and the filtrate was evaporated in vacuo. The residue was purified
by column chromatography on silica gel with ethyl acetate/
methanol (10:1, v/ v) as the elution solvent to afford 1.58 g of
2-hydroxy-1,2,3,6,7,11b-hexahydropyrido[2,1-a]isoquinolin-4-one (2)
as a yellow powder, followed by recrystallization from ether to
give colorless crystal: 1H NMR (CDCl3) δ 1.2-3.5 (6H, m), 4.2-
4.3 (1H, m, 2-H), 4.6-4.8 (2H, m, 3-H), 7.1-7.3 (4H, m, phenyl-
H); MS m/ z 217 ([M]+).
(7) Song, Q.; Greenway, G. M. Analyst 2 0 0 1 , 126, 37-40.
(8) Ridlen, J. S.; Skotty, D. R.; Kissinger, P. T.; Nieman, T. A. J. Chromatogr., B
1 9 9 7 , 694, 393-400.
(9) Skotty, D. R.; Nieman, T. A. J. Chromatogr., B 1 9 9 5 , 665, 27-36.
(10) Koike, K.; Li, Y.; Seo, M.; Sakurada, I.; Tezuka, K.; Uchikura, K. Biol. Pharm.
Bull. 2 0 0 0 , 23, 101-103.
(11) Monji, H.; Yamaguchi, M.; Aoki, I.; Ueno, H. J. Chromatogr., B 1 9 9 7 , 690,
305-313.
(12) Holeman, J. A.; Danielson, N. D. J. Chromatogr., A 1 9 9 4 , 679, 277-284.
(13) Uchikura, K.; Kirisawa, M. Anal. Sci. 1 9 9 1 , 7, 971-973.
(14) Jackson, W. A.; Bobbitt, D. R. Anal. Chim. Acta 1 9 9 4 , 285, 309-320.
(15) He, L.; Cox, K. A.; Danielson, N. D. Anal. Lett. 1 9 9 0 , 23 (2), 195-210.
(16) Targove, M. A.; Danielson, N. D. J. Chromatogr. Sci. 1 9 9 0 , 28, 505-509.
(17) Ridlen, J. S.; Klopf, G. J.; Nieman, T. A. Anal. Chim. Acta 1 9 9 7 , 341, 195-
204.
To a solution of 2 (1.40 g, 6.44 mmol) in 25 mL of tetrahydro-
furan in an ice bath was added LiAlH4 (25.8 mmol), and this was
refluxed for 2 h at 100 °C. After cooling, a small portion of
methanol and water was added to the reaction mixture to
decompose the excess reagent, followed by filtration, and the
filtrate was evaporated to dryness in vacuo. After addition of 10%
NaOH, the residue was extracted with chloroform. The organic
layer was dried with anhydrous MgSO4, and the filtrate was
(18) Downey, T. M.; Nieman, T. A. Anal. Chem. 1 9 9 2 , 64, 261-268.
(19) Martin, A. F.; Nieman, T. A. Anal. Chim. Acta 1 9 9 3 , 281, 475-481.
(20) Skotty, D. R.; Lee, W.-Y.; Nieman, T. A. Anal. Chem. 1 9 9 6 , 68, 1530-1535.
(21) Lee, W.-Y.; Nieman, T. A. J. Chromatogr., A 1 9 9 4 , 659, 111-118.
(22) Danielson, N. D.; He, L.; Noffisinger, J. B.; Trelli, L. J. Pharm. Biomed. Anal.
1 9 8 9 , 7, 1281-1285.
(23) Uchikura, K.; Kirisawa, M.; Sugii, A. Anal. Sci. 1 9 9 3 , 9, 121-123.
(24) Morita, H.; Konishi, M. Anal. Chem., 2 0 0 2 , 74, 1584-1589.
(25) Morita, H.; Konishi, M. J. Liq. Chromatogr. Relat. Technol. 2 0 0 2 , 25 (16),
2413-2423.
Analytical Chemistry, Vol. 75, No. 4, February 15, 2003 941

Figure 1. Schematic flow diagram of ECL detection.
evaporated in vacuo. The residue was purified by column chro-
matography on silica gel with chloroform/ methanol (35:1, v/ v)
as the elution solvent to afford 0.96 g of HPI (3 ) as a colorless
202 ([M]+); HR-MS m/ z calcd for C13H19N2 [M + H]+: 203.1548,
found 203.1557.
Synthesis of an HP I Derivative of Ibuprofen. After stirring
of a mixture of ibuprofen (10 mg, 48 µmol), DMF (1 drop), and
oxalyl chloride (48 µmol) in 1 mL of dichloromethane for 1 h,
HPI (48 µmol) and (dimethylamino)pyridine (96 µmol) were added
and the mixture was kept for 4 days at room temperature. After
addition of 5% NaHCO3, the reaction mixture was extracted with
chloroform. The organic layer was dried with anhydrous MgSO4,
and the filtrate was evaporated in vacuo. The residue was purified
by column chromatography on silica gel with n-hexane/ ethyl
acetate (9:2, v/ v) as the elution solvent to afford the ester: 1H
NMR (CDCl3) δ 0.88 (6H, d, J ) 4.4 Hz, (CH3)2CHCH2), 1.4-3.3
(11H, m), 1.50 (3H, d, J ) 4.8 Hz, CH(CH3)COO), 2.44 (2H, d, J
) 4.6 Hz, (CH3)2CHCH2), 3.70 (1H, q, J ) 5.0 Hz, CHCOO), 4.8-
5.0 (1H, m, 2-H), 7.0-7.3 (8H, m, phenyl-H); MS m/ z 391 ([M]+).
Synthesis of 3 -Isobutyl-9 ,1 0 -dimethoxy-1 ,3 ,4 ,6 ,7 ,1 1 b-
hexahydro-2 H -pyrido[2 ,1 -a ]isoquinolin-2 -ylam ine (IDH-
P IA). To a solution of tetrabenazine (500 mg, 1.58 mmol) in 30
mL of methanol were added ammonium acetate (15.8 mmol) and
sodium cyanoborohydride (1.1 mmol) followed by stirring for 22
h at room temperature. The reaction mixture was evaporated to
dryness in vacuo. The residue was dissolved with chloroform and
washed with 10% NaOH. The organic layer was dried with
anhydrous MgSO4 and the filtrate was evaporated in vacuo. The
residual oil was purified by column chromatography on silica gel
with chloroform/ methanol (40:1, v/ v) and alumina with ethyl
acetate as the elution solvent to afford 96 mg of IDHPIA as a white
1
crystal: mp 142-143 °C; H NMR (CDCl3) δ 1.4-3.2 (11H, m),
3.7-3.9 (1H, m, 2-H), 7.0-7.2 (4H, m, phenyl-H); MS m/ z 203
([M]+). Anal. Calcd for C13H17NO: C, 76.81; H, 8.43; N, 6.89.
Found: C, 76.52; H, 8.40; N, 7.00.
Synthesis of 1,3,4,6,7,11b-Hexahydro-2H-pyrido[2,1-a ]-
isoquinolin-2-ylamine (HP IA). Thionyl chloride (6.9 mmol) was
added dropwise to a solution of 3 (500 mg, 2.46 mmol) in 36 mL
of toluene in an ice bath, and the resultant mixture was stirred
for 1.5 h at 80 °C. After cooling, a small portion of water was added
to the reaction mixture to decompose the excess reagent. After
addition of 10% NaOH, the reaction mixture was extracted with
chloroform. The organic layer was dried with anhydrous MgSO4,
and the filtrate was evaporated in vacuo to afford 495 mg of
2-chloro-1,3,4,6,7,11b-hexahydro-2H-pyrido[2,1-a]isoquinoline (4 )
as a brown oil. Compound 4 was used without further purification.
To a solution of 4 (495 mg, 2.19 mmol) in 10 mL of
N,N-dimethylformamide (DMF) was added sodium azide (6.57
mmol) for 7 h at 100 °C. After evaporation, 10% NaOH was added
to the residue and the reaction mixture was extracted with
chloroform. The organic layer was dried with anhydrous MgSO4,
and the filtrate was evaporated in vacuo. The residue was purified
by column chromatography on silica gel with chloroform/
methanol (20:1, v/ v) and n-hexane/ ethyl acetate (1:1, v/ v) as the
elution solvent to afford 349 mg of 2-azido-1,3,4,6,7,11b-hexahydro-
2H-pyrido[2,1-a]isoquinoline (5 ) as a brown oil.
1
Compound 5 (99 mg, 0.43 mmol) was hydrogenated in ethanol
(3 mL) at room temperature and 1 atm in the presence of 5% Pd/ C
(30 mg) for 3.5 days. After filtration, the filtrate was evaporated
in vacuo. The residue was purified by column chromatography
on silica gel with chloroform/ methanol (10:1, v/ v) and n-hexane/
ethyl acetate (1:1, v/ v) as the elution solvent to afford 51 mg of
powder: mp 97-98 °C; H NMR (CDCl3) δ 0.91 (3H, d, J ) 6.0
Hz, (CH3)2CH), 0.94 (3H, d, J ) 6.0 Hz, (CH3)2CH), 1.0-3.4 (14H,
m), 3.84 (6H, s, 2 × CH3O), 6.57 (1H, s, phenyl-H), 6.68 (1H, s,
phenyl-H); MS m/ z 318 ([M]+); HR-MS m/ z calcd for C19H31N2O2
[M + H]+ 319.2386. found 319.2395.
Synthesis of IDHP IA Derivatives of P B and Myristic Acid
(C1 4 :0 ). To a solution of IDHPIA (2.0 mg, 6.28 µmol) in 0.3 mL of
acetonitrile were added PB (7.5 µmol), BEPT (12 µmol), and
1
HPIA (6 ) as a white powder: mp 199-200 °C (dec); H NMR
(CDCl3) δ 1.4-3.4 (12H, m), 7.1-7.3 (4H, m, phenyl-H); MS m/ z
942 Analytical Chemistry, Vol. 75, No. 4, February 15, 2003

MDPP (15 µmol) for 3.5 h at room temperature. The reaction
mixture was applied and purified by column chromatography on
silica gel with ethyl acetate as the developing solvent to afford
2.8 mg of the IDHPIA derivative of PB as a white powder: mp
186-187.5 °C; 1H NMR (CDCl3) δ 0.89 (3H, d, J ) 6.6 Hz, (CH3)2-
CH), 0.90 (3H, d, J ) 6.3 Hz, (CH3)2CH), 1.0-3.4 (20H, m), 3.79
(3H, s, CH3O), 3.84 (3H, s, CH3O), 6.59 (2H, s, phenyl-H), 7.2-
7.3 (5H, m, phenyl-H); MS m/ z 465 ([M + H]+).
The IDHPIA derivative of C14:0 was also synthesized in the
same manner and afforded 2.3 mg of the product as a white
powder: mp 47-48 °C; 1H NMR (CDCl3) δ 0.85-0.90 (9H, m, 3
× CH3), 1.1-3.2 (38H, m), 3.82 (3H, s, CH3O), 3.84 (3H, s, CH3O),
6.57 (2H, each s, phenyl-H); MS m/ z 528 ([M]+).
The detection pH was ∼2.1. The column was an Inertsil C8
(4.6 × 150 mm; GL Sciences Inc., Tokyo, Japan). The column
temperature was ambient (23 ( 2 °C). The photomultiplier applied
was biased at 0.7 kV.
RESULTS
Candidates for the Derivatization Reagent. In this system,
hydroxide ion reacts with Ru(bpy)33+; therefore, the background
level rises with an increase of pH. A lower detection pH leads to
a more stable analytical condition; therefore, the compounds that
can be detected at acid pH with strong intensity should serve as
good derivatization reagents. As most of the tertiary amines
showed maximum ECL intensity at alkaline pH, we searched for
suitable compounds. We focused on sparteine and dihydrocodeine,
which showed maximum ECL intensity at pH 3.0 and detection
limits of 50 and 2.5 fmol, respectively, at the signal-to-noise ratio
of 3. To develop more sensitive derivatization reagents, we
examined compounds closely related to sparteine and dihydro-
codeine. The results suggested that compounds fulfilling the
following criteria would show good signal-to-noise ratios: (i) rigid
amino compounds such as quinolizidine-derived compounds; (ii)
compounds containing a phenyl or mono- or dimethoxyphenyl
group; (iii) compounds showing high ECL intensity at acid pH
with high signal-to-noise ratio.
Optimization of Detection pH. The detection pH was
optimized by an FIA method. The carrier solution of 0.5 M
Britton-Robinson (BR) buffer (pH 1.5-9.0) and the reagent
solution of 0.8 mM Ru(bpy)3Cl2 in 10 mM H2SO4 were pumped at
the flow rates of 1.0 and 0.3 mL min-1, respectively. The BR buffer
was prepared as an acid solution with 61.3 g of boric acid, 56.5
mL of acetic acid, and 68 mL of phosphoric acid in 2 L of water.
The pH was adjusted with 2.5 M NaOH to prepare pH 1.5-9.0
buffers. Aliquots of 5 µL of 10 or 100% methanol solutions
containing 100 pmol of synthetic samples were injected triplicate,
and mean values were used throughout the study. Relative
standard deviations of most of the points were within 10% except
for the points that showed very weak ECL intensity. The
electrochemical oxidation was done by controlled-current mode
with the current maintained at 200 µA. The oxidation potential
was approximately +1.3 V versus R-hydrogen/ palladium reference
electrode. The photomultiplier applied was biased at 0.5 kV.
Derivatization P rocedure for P B. To a solution of 10 µM
PB in 50 µL of acetonitrile were added 50 µL each of 5 mM BEPT,
0.3 mM MDPP, and 1 mM IDHPIA solution in acetonitrile. After
mixing with a vortex mixer for a few seconds, the reaction mixture
was allowed to stand for 60 min at room temperature, and then
800 µL of 50% methanol containing 0.5% TFA was added to stop
the reaction. Of the resulting mixture, 10 µL was injected into
the chromatograph.
Selection of Derivatization Reagent and Optimization of
Detection pH. We evaluated the ECL intensity including the
background noise level defined as normalized ECL intensity (I),
which was calculated with the following equation: I ) A/ B, where
A is peak area per nanomole (µV‚s/ nmol) and B is background
current (nA).
We examined some amino compounds with quinolizidine or a
rigid structure. In the study of ECL intensity, we have found that
the compounds containing an alkoxyphenyl group tended to have
maximum ECL intensity at lower pH than the compounds without
an alkoxyphenyl group. No effect was observed with phenol-
derived compounds and these compounds showed optimum pH
at 7-9 (e.g., 9a-phenylquinolizidine). Quinolizidine-derived com-
pounds bonded with the phenyl group at a long distance also
showed no effect on optimum pH (e.g., 9a-(4-methoxybenzyl)-
quinolizidine) (data not shown). Consequently, we selected
benzoquinolizidine- and dimethoxybenzoquinolizidine-derived com-
pounds as candidates for derivatization reagents.
We synthesized HPI and HPIA as benzoquinolizidine-derived
compounds and IDHPIA as a dimethoxybenzoquinolizidin-derived
compound. The structure of benzoquinolizidines is shown in
Figure 2. Initially, HPI was synthesized by a one-step reduction
reaction with 1 in the presence of LiAlH4 but HPI was obtained
in low yield. HPI was synthesized by reducing 1 with NaBH4 to
give 2 , followed by reduction with LiAlH4. HPIA was synthesized
from HPI. IDHPIA was synthesized by a reductive amination
reaction of tetrabenazine with NaBH3CN. This reaction gave a
∼1:1 mixture of diastereoisomers. Although we examined the
reaction with dimethylamine borane, which is a reductive amina-
tion reagent reacting by different mechanisms with NaBH3CN,
the proportion of the two products did not change. We used the
less polar compound of the diastereoisomers as the derivatization
reagent.
Derivatization P rocedure for Fatty Acids in Human Se-
rum. Blood samples were obtained from healthy human volun-
teers (male, aged from 21 to 51 years) in our laboratories after
overnight fasting. A 5-µL aliquot of human serum was mixed with
200 µL of 0.5 M phosphate buffer (pH 6.5), 50 µL of the 0.4 µM
margaric acid (C17:0, internal standard) solution in acetonitrile, and
2.0 mL of a mixture of chloroform/ n-hexane (1:1, v/ v). The
mixture was vortexed for ∼5 min and centrifuged at 2000 rpm
for 5 min, and the organic layer was evaporated under nitrogen
stream. The residue was dissolved in 50 µL of acetonitrile, and
then 50 µL each of the 5 mM BEPT, 1 mM MDPP, and 1 mM
IDHPIA solution in acetonitrile was added. After mixing with a
vortex mixer for a few seconds, the reaction mixture was allowed
to stand for 45 min at room temperature, and then 800 µL of 80%
acetonitrile was added to stop the reaction. Of the resulting
mixture, 10 µL was injected into the chromatograph.
HP LC Conditions for Fatty Acids. The mobile phase of
52% acetonitrile containing 0.05% trifluoroacetic acid (TFA) and
reagent solution of 0.8 mM Ru(bpy)3Cl2 in 10 mM H2SO4 were
delivered at the flow rates of 1.0 and 0.3 mL min-1, respectively.
The pH-dependent ECL intensity was examined for benzo-
quinolizidine-derived compounds. As shown in Figure 3, benzo-
Analytical Chemistry, Vol. 75, No. 4, February 15, 2003 943

Figure 2. Structure of benzoquinolizidine-derived compounds.
Figure 4. pH profile of HPI derivative of ibuprofen and IDHPIA
derivatives of PB and C14:0: O, IDHPIA derivative of PB; 4, C14:0; 0,
HPI derivative of ibuprofen. Samples: 100 pmol in 10 or 100%
methanol solution. Each point was measured in triplicate, and
standard deviation is indicated as error bars.
Figure 3. pH profile of benzoquinolizidine-derived compounds: O,
tetrabenazine; 0, IDHPIA; 4, HPI; 2, HPIA; b, N,N-diisopropylethyl-
amine. Samples: 100 pmol in 10% methanol solution. Each point
was measured in triplicate, and standard deviation is indicated as
error bars.
quinolizidine-derived compounds such as HPI and HPIA showed
the maximum normalized ECL intensity at pH 5.0. The values of
the normalized ECL intensity increased 10-41-fold for dimethoxy-
benzoquinolizidine-derived compounds such as tetrabenazine and
IDHPIA compared to the benzoquinolizidine-derived compounds
and showed maximum intensities at pH 1.5-2.0 in the tested
range. These compounds gave 3-120-fold higher normalized ECL
intensities than N,N-diisopropylethylamine, which is one of the
derivatives of ibuprofen were both 15 fmol with the proposed
HPLC separation (data not shown). IDHPIA was more sensitive
than HPI and HPIA. IDHPIA was used for the following experi-
ment.
Optimization of the Derivatization Reaction. For the de-
rivatization of carboxylic acids, BEPT and MDPP were selected
as the condensation reagent and the acid-capturing reagent for
HBr and HBF4 formed in the reaction, respectively. These
reagents allowed the reaction to proceed with carboxylic acids to
produce the corresponding amides at room temperature within 1
h (Figure 5), and electrogenerated chemiluninescence is based
3+
most responsive compounds with Ru(bpy)3 and showed a
detection limit of 50 fmol at the signal-to-noise ratio of 3
(unpublished data).
2+
The pH-dependent normalized ECL intensity was examined
for HPI, HPIA, and IDHPIA derivatives of some carboxylic acid
compounds with the synthetic samples. These derivatives showed
the same optimum pH as the unbound derivatization reagents as
shown in Figure 4. The normalized ECL intensities of IDHPIA
derivatives of PB and C14:0 were 153 and 50% relative to that of
unbound IDHPIA, respectively. The normalized ECL intensity of
the HPI derivative of ibuprofen was 83% relative to that of unbound
HPI. Only a small effect was observed for the intensity with
condensation of the analyte; therefore, these results show that a
wide range of compounds could be used without significant
decrease of ECL intensity. The detection limits of HPI and HPIA
on the reaction between Ru(bpy)3 and tertiary amine of the
benzoquinolizidine-derived compounds. Moreover, chromato-
graphic separation of the reagent and the conjugate was easy and
no interference was observed for the analysis of the analyte. C14:0
was used as a representative fatty acid to optimize the reaction
condition. The effect of the concentration of IDHPIA, BEPT, and
MDPP on the reactivity to C14:0 was examined at 0-5 mM. The
maximum peak area was obtained at 1 mM for IDHPIA and MDPP
(50 equiv) and 5 mM for BEPT (250 equiv). Figure 6 shows the
time course of the reaction under these conditions. A quantitative
reaction was observed at more than 45 min. The derivatization
condition for PB was also optimized at 1 mM IDHPIA (100 equiv),
944 Analytical Chemistry, Vol. 75, No. 4, February 15, 2003

Figure 5. Reaction scheme of fatty acids with IDHPIA.
Table 2. Concentration of Fatty Acids in Human Serum
(nmol mL-1
)
subject C12:0 C14:1 C14:0 C18:3 C16:1 C18:2 C16:0 C18:1 C18:0
1
2
3
4
5
6
7
1.2 0.6
8.1 0.6
3.8 0.7 12.7
0.3 1.8
0.2 0.8
0.3 0.8
0.1 1.1
4.1 15.3
8.4 7.8
3.8 89.0
2.8 78.6
88.4
85.3
89.5
63.9
83.0 22.5
88.8 24.2
93.1 21.4
90.4 13.5
6.3 12.0 48.5
9.3 12.5 62.1
9.8 11.6 61.0
5.3 10.4 71.9
1.4
2.5
1.2
3.1
98.0 141.8 33.6
90.2 133.1 23.1
7.9 17.6 76.2 107.3 159.0 27.7
mean
SD
2.0 0.9
3.0 0.4
4.8
4.3
8.8 10.1 69.6 88.9 112.7 23.7
3.3
4.4
5.2 13.4
4.7 59.4
13.3
91.9
30.9
6.1
meana 1.5 1.5
(SD)
7.9
90.3 21.2
(1.8) (1.1) (4.9) (2.4) (1.5) (23.4) (36.3) (42.1) (9.1)
meanb 5.8 ndc nd
3.9 18.5 52.1 80.7 93.7 31.5
(1.1) (1.9) (15.5) (26.4) (38.9) (7.8)
Figure 6. Time course of the derivatization reaction of IDHPIA with
(SD)
(0.6)
C
14:0 in the presence of BEPT and MDPP. Conditions: 20 µM C14:0,
5 mM BEPT, 1 mM MDPP, and 1 mM IDHPIA in acetonitrile, room
temperature. Each point was measured in duplicate.
a
Quoted from ref 26. b Quoted from ref 27. c nd, not determined.
Table 1. Recovery of Fatty Acids (n ) 5)
compound
recovery (%)
compound
recovery (%)
interfere with the analyte peaks. Under these conditions, the
detection pH was ∼2.1. The optimum detection pH was the same
as that of the mobile phase; therefore, no pump was needed to
deliver pH conditioning. Nieman et al. reported a versatile Ru-
C12:0
C14:1
C14:0
C18:3
C16:1
98.8 ( 6.1
93.7 ( 5.0
102.0 ( 7.2
94.9 ( 5.8
96.1 ( 6.0
C18:2
C20:4
C16:0
C18:1
C18:0
94.7 ( 5.8
94.0 ( 4.9
105.0 ( 6.5
94.6 ( 5.6
97.7 ( 6.1
2+
(bpy)3 ECL detection technique with the CL reagents added
directly to the mobile phase.8,17,20 This technique requires just one
pump to deliver the mobile phase containing the CL reagent. We
examined the technique to try to develop more sensitive detection
and simpler equipment. But more sensitive detection could not
be achieved in our system. This result would be caused by the
difference of the flow cell. They used their own apparatus with a
laminar-type flow cell with a cell volume of 9.2 µL, and the potential
was applied at the flow cell. Our system contains a spiral flow
cell with a cell volume of 80 µL, and the potential was applied
before the cell inlet.
5 mM BEPT (500 equiv), and 0.3 mM MDPP (30 equiv) for 60
min.
Determination of Fatty Acids in Human Serum. Lauric acid
(C12:0), myristoleic acid (C14:1), myristic acid (C14:0), R-linolenic acid
(C18:3), palmitoleic acid (C16:1), arachidonic acid (C20:4), R-linoleic
acid(C18:2), palmitic acid (C16:0), oleic acid (C18:1), and stearic acid
(C18:0) were used as free fatty acids and C17:0 was used as the
internal standard. Free fatty acids were extracted with a mixture
of chloroform and n-hexane (1:1, v/ v). The recovery of all fatty
acids tested was over 93% as shown in Table 1. The photomultiplier
voltage was optimized for IDHPIA derivatives of fatty acids, and
a bias of 0.7 kV was used for the determination of fatty acids.
Figure 7 shows typical chromatograms of fatty acids in standard
solution and extracted samples from human serum. Separation
of the 11 fatty acid derivatives was satisfactorily achieved on the
reversed-phase column by isocratic elution mode with the 0.05%
TFA/ acetonitrile. Each peak shape of IDHPIA derivatives of fatty
acids was good and the theoretical numbers were between 29 000/
m (C12:0) and 59 500/ m (C18:0). Each analyte was identified by
comparison of its retention time to that of the fatty acid standard.
All the fatty acid derivatives showed the same peak area; therefore,
differences of ECL intensity between the compounds would be
small. The excess reagent eluted earlier at the void and did not
The detection limits of the HPLC analysis were found to be
0.6 and 0.5 fmol, respectively, for PB and C14:0 at the signal-to-
noise ratio of 3. Standard solutions of different concentrations of
PB, C12:0, and other nine fatty acids ranging from 0.25 to 500 (n )
12), 0.5 to 1000 (n ) 8), and 2 to 2000 nM (n ) 8) were derivatized,
respectively. All the calibration curves were linear for the range
tested (correlation coefficients >0.999).
Table 2 shows the concentrations of free fatty acids in sera
from healthy volunteers determined by this method. The mean
values for the individual free fatty acids in normal serum were in
good agreement with the published data.26,27
(26) Yamaguchi, M.; Matsunaga, R.; Hara, S.; Nakamura M.; Ohkura, Y. J.
Chromatogr., B 1 9 8 6 , 375, 27-35.
(27) Kargas, G.; Rudy, T.; Spennetta, T.; Takayama, K.; Querishi, N.; Shrago, E.
J. Chromatogr., B 1 9 9 0 , 526, 331-340.
Analytical Chemistry, Vol. 75, No. 4, February 15, 2003 945

Figure 7. Chromatograms of IDHPIA derivatives of fatty acids of (a) standard solution (2 pmol each on column) and (b) extracts from human
serum. HPLC conditions are given in the Experimental Section. Peaks: 1, C12:0; 2, C14:1; 3, C14:0; 4, C18:3; 5, C16:1; 6, C18:2; 7, C20:4; 8, C16:0; 9,
C18:1; 10, C17:0; 11, C18:0.
CONCLUSIONS
wide range of drugs and endogenous compounds in biological
fluids.
IDHPIA, a new electrogenerated chemiluminescent derivati-
zation reagent for carboxylic acids, was specifically developed
for the ECL system. The IDHPIA derivative showed optimum
pH in the acid region and was 100-fold more sensitive than
reagents reported previously. Also, because the optimum pH was
the same for the mobile phase, the equipment could be simplified.
This reagent could be synthesized with a one-step reaction from
the commercially available compound and should be useful for a
ACKNOWLEDGMENT
We thank Ms. R. Nishimura and Ms. Y. Nakano for measuring
elemental analyses and high-resolution mass spectra, respectively.
Received for review June 5, 2002. Accepted December 12,
2002.
AC020377I
946 Analytical Chemistry, Vol. 75, No. 4, February 15, 2003