3704 M. N. Uddin et al. J. Sep. Sci.

2008, 31, 3704 – 3717

Mohammad Nasir Uddin Original Paper

Victoria F. Samanidou
Ioannis N. Papadoyannis
Laboratory of Analytical
Chemistry, Department of
Chemistry, Aristotle University of
Thessaloniki, Thessaloniki,
Greece
Validation of SPE-HPLC determination of
1,4-benzodiazepines and metabolites in blood
plasma, urine, and saliva
A simple, sensitive, selective, and reproducible RP-HPLC method with DAD detection
at 240 nm was developed for the determination of six 1,4-benzodiazepines: brom-
azepam (BRZ), clonazepam (CLZ), diazepam (DZP), flunitrazepam (FNZ), lorazepam
(LRZ), alprazolam (APZ); and two metabolites: a-hydroxyalprazolam (HALZ) and a-
hydroxytriazolam (HTZL) in human plasma, urine, and saliva, using colchicine as
internal standard, after SPE using Nexus Varian cartridges. Separation was perform-
ed on a Kromasil C8 (250 mm65 mm, 5 lm) analytical column with a gradient
mobile phase containing methanol, ACN and 0.05 M ammonium acetate. Linearity
was held within the range 0.3 – 20.0 ng/lL, with coefficients of determination (r2)
better than 0.997. The within- and between-day assay RSD at 2, 4, 8 ng/lL ranged
from 0.03 to 4.7% and 0.5 to 7.0%, respectively in standards, from 1.3 to 7.9% and
3.3 to 7.3%, respectively in plasma, from 2.1 to 6.0% and 2.1 to 7.8%, respectively in
urine and at 0.5, 1.0, 2.0 ng/lL ranged from 2.22 to 5.8% and 2.2 to 8.1%, respec-
tively, in saliva. The mean relative recoveries were 96.3 – 108.6, 96.0 – 108.2, 94.3 –
107.1, 97.0 – 107.0% in within-day assay and 96.8 – 107.7, 94.6 – 107.6, 93.2 – 105.8,
96.0 – 108.6 in between-day assay for standard, plasma, urine, and saliva, respec-
tively. The LOD and LOQ were 0.02 – 0.47 and 0.07 – 1.57 ng/lL, respectively.
Keywords: Benzodiazepines / Blood plasma / High-performance liquid chromatography / Metabo-
lites / Solid-phase extraction / Urine and saliva /
Received: June 12, 2008; revised: August 18, 2008; accepted: August 19, 2008
DOI 10.1002/jssc.200800342

1 Introduction also frequently encountered in clinical and forensic case-

Benzodiazepines (BDZs) constitute a large and important
class of pharmaceutical active drugs with antiepileptic
hypnotic, tranquillizing, anticonvulsant, sedative,
muscle relaxant, and amnesic properties. Thirty-five of
them are controlled by the United Nations convention
(1971), and more than 50 are available throughout the
world [1 – 3]. Since the introduction of the first 1,4-benzo-
diazepine, chlordiazepoxide [3, 4], in 1960, they have
become the drugs of choice for the treatment of anxiety,
sleep disorders, status epilepsy, insomnia, frequent noc-
turnal or early morning awakenings and other convul-
sive disorders [5, 6]. Because of these properties they are
Correspondence: Professor Ioannis N. Papadoyannis, Laboratory
of Analytical Chemistry, Department of Chemistry, Aristotle
University of Thessaloniki, Thessaloniki, GR-541 24, Greece
E-mail: papadoya@chem.auth.gr
Fax: +30-2310997719
Abbreviations: APZ, alprazolam; BDZ, benzodiazepine; BRZ,
bromazepam; CLZ, clonazepam; DAD, diode array detector;
DZP, diazepam; FNZ, flunitrazepam; HALZ, a-hydroxyalprazo-
lam; HTZL, a-hydroxytriazolam; LRZ, lorazepam
work samples involving road traffic offences and/or drug
overdoses [2].
Several analytical methods have been published for
the quantification of BDZs and/or their metabolite(s) in
human (or animal) biological fluids or tissue organ
extracts based on acid – base titration [7], polarography
[8], potentiometry [9], fluorimetry [10], spectrophotome-
try [11], MEKC (MECC) [12], adsorptive-stripping voltam-
metry [13], differential pulse voltammetry (DPV) [14], TLC
[15], immunoassay [16], RIA [17], and flow injection anal-
ysis (FIA) [18]. By far most of the published methods use-
ful in clinical and forensic laboratories are applying
chromatography. These include GC with electron cap-
ture detection (GC-ECD) [19, 20], nitrogen – phosphorus
detection (GC-NPD) [20, 21], MS with negative ion chemi-
cal ionization (GC-MS/NICI) [22] or SIM (GC-MS/SIM) [23],
GC/MS/MS [24], and HPLC methods with several detection
modes including UV (HPLC-UV) [25] or diode array detec-
tor (HPLC-DAD) [26, 27], mass spectrometric detection
[28] or HPLC/MS/MS [29]. HPLC/LC-MS or MS/MS interfaced
with ESI or atmospheric pressure chemical ionization
(APCI) further improves the specificity and sensitivity of

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2008, 31, 3704 – 3717
detection and has emerged as a selective analytical
method in drug analysis [30 – 33].
However, each of these methods has its shortcomings
such as lack of selectivity, inadequate sensitivity, use of
large sample volumes (F 1 mL) or use of expensive SPE
cartridges. GC-MS is an excellent method that provides
unambiguous identification of compounds with good
sensitivity; but needs an expensive instrument, time-con-
suming sample clean-up thus increasing the complexity;
suffers from the thermal instability at high column tem-
perature used and low volatility of many BDZs requiring
a derivatization step [33]. Though the combination of MS
offers the advantage of the separation power of HPLC
with its sensitivity and specificity, the technique involves
expensive equipment, which is not affordable for most
nonresearch laboratories, particularly those in resource-
poor countries.
Again, an effective sample pretreatment procedure,
which removes interfering endogenous components and
preconcentrates the analytical compounds prior to HPLC
separation is often necessary to improve sensitivity.
Either liquid – liquid extraction [34], SPE [26, 27, 30], on-
line SPE [32], or solid-phase microextraction [35] was usu-
ally employed, which could enrich the analytes by sev-
eral folds, even 1 – 2 orders of magnitude. An alternative
sample pretreatment method involves protein removal
by precipitation (with organic solvent) and centrifuga-
tion. The isolated supernatants could then be directly
introduced into HPLC [36, 37] for determination. Owing
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Liquid Chromatography 3705
Figure 1. Chemical structures of
BDZs and metabolites.
to lack of a preconcentration step, the direct supernatant
introduction technique could not provide the sensitivity
high enough for pharmacokinetic study of the drug.
However, SPE is still being exploited by the researchers
for attaining objectives, like reduced solvent consump-
tion and disposal, reduced time, cost, and labor provid-
ing higher concentration factors than LLE. It also pro-
vides greater opportunity for selective isolation, such as
fractionation of the sample into different compounds or
groups of compounds [38]. To the best of our knowledge
only one paper has reported [39] applying SPE for the iso-
lation of BDZs from saliva.
The present paper describes the development and vali-
dation of a relatively simple, sensitive, and selective
HPLC-DAD method for the determination of six 1,4-ben-
zodiazepines including two metabolites as shown in Fig.
1, in human plasma, urine, and saliva samples. The gra-
dient elution provided a good separation of analytes. SPE
was used for sample pretreatment yielding extracts free
from endogenous interference.
2 Experimental
2.1 Instrumentation
A Shimadzu (Kyoto, Japan) quaternary low-pressure gra-
dient system was used for the chromatographic determi-
nation of the examined analytes. The solvent lines were
mixed in an FCV-10ALVP mixer. An LC-10ADVP pump
www.jss-journal.com
3706 M. N. Uddin et al.
equipped with a Shimadzu SCL-10ALVP System Control-
ler, permitting fully automated operation, was used to
deliver the mobile phase to the analytical column. Sam-
ple injection was performed via a Rheodyne 7725i injec-
tion valve (Rheodyne, Cotati, California, USA) equipped
with a 20 lL loop. Detection was achieved by an SPD-
M10AVP Photodiode Array Detector, complied with data
acquisition software Lab Solutions-LC-Solutions by Shi-
madzu. Degassing of the mobile phase was performed by
helium sparging in the solvent reservoirs by a DGU-10B
degassing unit.
The analytical column, a Kromasil C8 (250 mm64
mm, 5 lm) was purchased from MZ-Analysentechnik
(Mainz, Germany). A glass vacuum-filtration apparatus
obtained from Alltech Associates was employed for the
filtration of the buffer solution, using 0.2 lm cellulose
nitrate 0.2 lm-WCN Type (47 mm DIA) membrane filters
(Whatman Laboratory Division, Maidstone, England).
The SPE study was carried out on a 12-port vacuum mani-
fold from Supelco (Bellefonte, PA, USA). LC-18 cartridges
500 mg/3 mL and DSC-18 500 mg/3 mL were supplied
from Supelco, Abselut Nexus (30 mg/1 mL) from Varian
and Lichrolut RP-select B (200 mg/3 mL) from Merck. All
evaporations were performed with a Supelco 6-port Mini-
Vap concentrator/evaporator.
2.2 Chemicals and reagents
Colchicine (95% HPLC) used as the internal standard was
supplied by Sigma (Sigma – Aldrich Chemie BV, The Neth-
erlands). HPLC-grade organic solvents (methanol, ACN)
were supplied by Carlo Erba (Milano, Italy). Water used
throughout the study was purified as required by the
reverse osmosis method to gain high-purity water with a
Milli-Q water purification system from Millipore (Milli-
pore, Bedford, MA, USA). Ammonium acetate p.a. was
supplied from Riedel-de Haen (Buchs, SG, Switzerland).
Six drug standards and two metabolites were from Sigma
Chemical Company (St. Louis, MO, USA) and Biomol (Ger-
many), respectively. Purity of reference compounds was
not less than 98%.
2.3 Collection of specimens
Drug free plasma samples were kindly provided from the
Blood Donation Unit of a State Hospital, while urine and
saliva samples were provided by one male healthy volun-
teer. Pooled samples were prepared. Both urine and sal-
iva samples were collected in polypropylene tube. Saliva
samples were collected over 180 min by spitting at fixed
time intervals, without any stimulation of the saliva pro-
duction, to obtain approximately 5 mL of sample. By the
30 min period before sampling no food or drink was
taken and the mouth was rinsed with fresh water before
spitting [25, 26]. Supplied drug-free plasma and as soon
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2008, 31, 3704 – 3717
Table 1. Gradient program for chromatographic conditions of
proposed method
Time Mobile phase composition (%)
CH3OH CH3CN 0.05 M CH3COONH4
0.01 14 32 54
9.00 13 32 55
11.00 11 28 61
12.00 40 28 32
as possible after collection without any preservative the
urine and the supernatant of saliva obtained after centri-
fugation at 2500 g for 10 min to remove potential food
debris, were separated into 1.5 mL aliquots in eppendorf
and stored at – 208C until analysis (performed within
15 days).
2.4 Chromatographic conditions
Chromatography was performed under gradient condi-
tions at ambient temperature (about 258C). The mobile
phase consisted of a mixture of methanol, ACN, and
ammonium acetate buffer (0.05 M) at initial composition
of 14:32:54 by volume delivered in gradient mode as
shown in Table 1 at a flow rate of 1.1 mL/min, generating
an operating backpressure of initially 230 bar, which
remained unchanged up to 14 min and finally reduced
to 180 bar at 16 min. Before use, the mobile phase was
degassed for 10 min with helium. The column effluent
was monitored with a variable wavelength UV – Vis detec-
tor set at 240 nm.
2.5 Standard stock solutions
Stock solutions (100 ng/lL) of BDZs, metabolites and col-
chicine (IS) were prepared by dissolving an appropriate
amount of each compound in methanol and were stored
at 48C, protected from light and used within 3 months.
The stock solutions of drugs were further diluted daily
before analysis with methanol to make seven working
mixture solutions (controlled solution) at concentrations
of 0.20 – 20.00 lg/mL for each compound containing
internal standard at a constant concentration of 4 ng/lL.
A 0.05 M aqueous solution of ammonium acetate was
prepared by mixing the appropriate weight in Milli-Q
water and filtered before use.
2.6 Optimization of SPE protocol
SPE protocols were optimized in terms of recovery stud-
ies of BDZs prior to the application to biological fluids.
Various extraction protocols were tested to isolate the
analytes from samples using different sorbents or mix-
www.jss-journal.com
J. Sep. Sci. 2008, 31, 3704 – 3717 Liquid Chromatography 3707

Table 2. Study of column efficiency

Analytes Retention Separation Resolution No. of theoretical Asymmetry

factor factor factor plates factor
k = (tR – t0) a = k2 – k1 Rs = 2(tR2 – tR1)/(w1 + w2) N = 16(tR/w)2 AsF = b0.1/a0.11

IS 1.1 – – 1 228.2 1.8

BRZ 2.0 1.8 3.2 1 577.0 1.8
HTRZ 2.9 1.4 3.5 5 449.1 1.5
HALZ 3.1 1.1 1.2 6 171.4 1.3
CLZ 3.8 1.2 3.2 6 501.3 1.0
FNZ 4.1 1.1 1.2 7 324.9 1.0
LRZ 4.4 1.1 1.3 8 246.6 1.0
ALZ 4.8 1.1 1.3 9 188.4 1.0
DZP 7.5 1.6 13.1 35 656.3 1.5

ture of solvents at different compositions for elution of 2.8 Method validation

the adsorbed analytes. Extraction yield was estimated by
the comparison of peak area ratios of extracted standard
solutions versus nonextracted solutions.
2.7 Sample preparation
2.7.1 Preparation of spiked samples
Experimental standard stock solution (200 lL) at seven
concentration levels of 0.2 – 20 ng/lL for spiked samples
or 200 lL of methanol for blank samples was added to ali-
quots of 50 lL of pooled plasma or 100 lL of urine or
500 lL saliva samples.
2.7.2 Sample pretreatment
Prior to extraction, biological fluids were pretreated for
the removal of proteins by precipitation with the addi-
tion of 200 lL of ACN to spiked samples; to aliquots of
50 lL of pooled blood plasma or 100 lL of urine or 500 lL
saliva samples (if necessary) or spiked blank samples.
After centrifugation for 15 min at 3000 rpm, the clear
supernatant was applied to SPE cartridges.
2.7.3 SPE
Prior to extraction SPE cartridges were conditioned with
2 mL of methanol and 2 mL of water. Subsequently
spiked samples were applied after addition of 2 mL dis-
tilled water to decrease the percentage content of ACN of
supernatants. After washing with 1 mL water retained
drugs were eluted with (261) mL MeOH/CH3CN (50:50)
followed by the evaporation to dryness under a gentle
flow of N2 at 308C. The residue was reconstituted with
200 lL of methanol prior to their injection into the
liquid chromatographic system or the tubes were closed
before storing at –208C. Aliquots of 20 lL were injected
onto column.
The developed method was validated in terms of ICH ana-
lytical performance parameters like (CPMP/ICH/281/95,
Note for guidance on validation of analytical procedures:
Methodology, ICH topic Q2B validation of analytical pro-
cedures: Methodology, Step 4 (CPMP Adopted December
96)) precision, accuracy, selectivity, LOD, LOQ, linearity,
system suitability, robustness, and stability.
3 Results and discussion
3.1 Chromatography
An RP HPLC procedure with DAD detection has been
described for the selective, sensitive, accurate, and repro-
ducible quantitative analysis of BDZs and two metabo-
lites in human biological samples. In compromising
peak resolution and retention time, the mentioned con-
ditions were accepted when the examined analytes were
well resolved providing better separation. The separation
of BDZs and the internal standard was achieved in less
than 16 min. Retention times of the examined BDZs are
3.92 l 0.06 for the internal standard, 5.47 l 0.06 for
bromazepam (BRZ), 7.17 l 0.11 for a-hydroxytriazolam
(HTZL), 7.66 l 0.11 for hydroxyalprazolam (HALZ),
9.00 l 0.13 for clonazepam (CLZ), 9.55 l 0.15 for flunitra-
zepam (FNZ), 10.17 l 0.0.19 for lorazepam (LRZ),
10.77 l 0.21 for ALP, and 15.77 l 0.12 min for diazepam
(DZP). A representative chromatogram of extracted blank
and spiked human plasma, urine, and saliva, obtained
under the described chromatographic conditions are
illustrated in Figs. 2 – 4.
3.2 Column efficiency
Retention factor (k = tR – t0/t0), separation factor (a = tR2 –
t0/tR1 – t0), resolution factor Rs = 2(tR2 – tR1)/(w2 + w2), num-

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

3708 M. N. Uddin et al. J. Sep. Sci. 2008, 31, 3704 – 3717

Figure 2. Chromatogram of (A) blank plasma and (B) plasma spiked with BDZs (4 ng/lL) under the conditions described in text.
Peaks are as follows: IS (3.942 min), BRZ (5.551 min), HTRZ (7.237 min), HALZ (7.698 min), CLZ (9.081 min), FNZ (9.622 min),
LRZ (10.173 min), ALZ (10.800 min), and DZP (15.759 min).

ber of theoretical plates N = 16(tR/w)2 and asymmetry fac- times and w1 and w2 the baseline peak widths of succes-
tor (AsF = b0.1/a0.1) have been calculated to check the col- sive peaks. Results given in Table 2 show the excellent
umn efficiency. Here, t0, tR1, and tR2 are the retention performance of the column.

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2008, 31, 3704 – 3717 Liquid Chromatography 3709

Figure 3. Chromatogram of (A) blank urine and (B) urine spiked with BDZs (8 ng/lL) under the conditions described in text,
where peak sequence is as in Fig. 2.

3.3 Sample preparation and extraction

different compositions have been tested to optimize the
The sample preparation procedure used in this study SPE protocol by means of the recovery of analytes com-
involved only a single step, i.e., SPE. Four different sorb- pared to that of unextracted samples. From the results
ents and two organic solvents (methanol and ACN) in presented in Tables 3 and 4, Nexus Varian and 50:50

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

3710 M. N. Uddin et al. J. Sep. Sci. 2008, 31, 3704 – 3717

Figure 4. Chromatogram of (A) blank saliva and (B) saliva spiked with BDZs (4 ng/lL) under the conditions described in text,
where peak sequence is as in Fig. 2.

MeOH/ACN yielded the best recoveries for all analytes. of nonextracted analyte)6100 (%)). Optimization of sam-
The results were expressed as percentage recovery (i.e., ple pretreatment involved the result of a clean chromato-
recovery (%) = (peak area of extracted analyte/peak area gram as well.

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2008, 31, 3704 – 3717 Liquid Chromatography 3711

Table 3. Optimization of SPE sorbent for the isolation of ana- 3.5.2 LOD and LOQ
lytes
The LOD, defined as the lowest absolute amount of ana-
Analytes Recovery (%) lyte in a sample can be detected but not necessarily quan-
tified and the LOQ, the lowest amount of analyte in a
Lichrolut Varian LC-18 DSC-18 sample that can be determined with acceptable precision
and accuracy (i.e., f 20% (n = 6)) were calculated from
BRZ 92.0 96.7 99.0 91.6
HTRZ 98.5 104.9 94.5 89.0 regression equation using the formula: LOD = 3Sxy/a, and
HALZ 97.7 105.9 95.0 91.7 the LOQ: LOQ = 10Sxy/a, where Sxy is the SD of intercept
CLZ 102.4 98.2 102.7 92.4 and a is the slope [40].
FNZ 94.3 103.9 96.8 91.3 LOD values were 0.02 – 0.15, 0.08 – 0.47, 0.08 – 0.34 ng/
LRZ 99.8 97.5 98.4 93.0
ALZ 106.1 98.4 108.0 87.6 lL, 0.04 – 0.14 for standard, plasma, urine, and saliva,
DZP 93.3 105.3 103.9 91.8 respectively and corresponding LOQ values were 0.07 –
0.53, 0.22 – 1.57, 0.26 – 1.59 ng/lL and 0.13 – 0.45 ng/lL.

3.4 Preparation of calibration curves 3.5.3 Accuracy/recovery

Calibration curves were prepared for seven concentra-
tion levels ranging from 0.2 to 20.0 ng/lL of analytes for
standard and spiked bio-fluids (human plasma, urine,
and saliva). The biological samples were analyzed accord-
ing to the extraction procedure described in Section 2.
The calibration curves were constructed by plotting peak
area ratios of corresponding analytes to the internal
standard against their theoretical concentrations which
were fitted by a least squares linear regression to the
equation: response ratio (y) = slope6concentration
(x) + intercept. The regression coefficient (r2) of the cali-
bration lines, the day-to-day standard error in the slopes
and the intercepts are given in Table 5.
3.5 Method validation
3.5.1 Linearity
Calibration curve is linear over the range 0.2 – 20 ng/lL
studied in standard solutions for all analytes except
alprazolam (APZ). For APZ in standard solutions as well
as for all analytes in spiked plasma, urine, and saliva sam-
ples linearity holds from 0.3 to 20.0 ng/lL, with coeffi-
cients of determination (r2) better than 0.997.
The accuracy of the method is defined as the degree of
agreement of test results generated by the method to the
true value. The analytical recovery (extraction efficiency)
was used to assess the accuracy and it was measured by
spiking drug-free plasma, urine, and saliva samples with
known concentrations of the standards (QC samples).
The QC samples at three different concentration levels
(2.0, 4.0, 8.0 ng/lL for standard, plasma, urine, and 0.05,
1.0, 2.0 ng/lL for saliva) were assessed for the evaluation
of the accuracy of the method. The chromatographic
peak area ratios of analytes extracted from spiked
plasma, urine, and saliva at each level of the QC samples
to the area ratio of IS were used to determine the concen-
tration using respective calibration equation and it was
compared with the corresponding nominal concentra-
tion. The results were expressed as percentage recovery
(i.e., recovery (%) = (calculated concentration/nominal
concentration)6100 (%)).
The mean relative recoveries at three concentration
levels shown in Tables 6 – 9 for all analytes were 96.3 –
108.6, 96.0 – 108.2, 94.3 – 107.1, 97.0 – 107.0% in within-
day analysis and 96.8 – 107.7, 94.6 – 107.6, 93.2 – 105.8,
96.0 – 108.6% in between-day analysis for standard,
plasma, urine, and saliva, respectively.

Table 4. Optimization of elution solvents for SPE using Nexus Varian sorbent

Eluent composition Recovery (%)

BRZ HTRZ HALZ CLZ FNZ LRZ ALZ DZP

MeOH/ACN = 100:0 98.0 100.1 99.2 102.4 105.8 96.4 98.8 99.8
MeOH/ACN = 80:20 96.8 104.0 103.7 105.0 96.8 98.4 95.8 99.3
MeOH/ACN = 60:40 93.1 100.8 106.3 104.0 96.1 96.4 94.9 94.4
MeOH/ACN = 50:50 96.3 104.1 102.9 102.8 102.7 99.3 96.2 100.9
MeOH/ACN = 40:60 80.3 93.4 97.6 94.7 90.6 96.7 85.4 93.9
MeOH/ACN = 20:80 60.7 84.3 81.7 86.6 80.7 77.4 78.1 85.2
MeOH/ACN = 0:100 20.4 51.7 30.9 90.5 71.0 6.5 88.7 89.6

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

3712 M. N. Uddin et al. J. Sep. Sci. 2008, 31, 3704 – 3717

Table 5. Sensitivity and linearity data of the proposed method

Analytes Regression equation r2 LOD LOQ Linearity range

(ng/lL) (ng/lL) (ng/lL)

Standard
BRZ y = (0.307 l 0.002)x + (0.021 l 0.002) 0.999 0.02 0.07 0.2 – 20
HTRZ y = (0.195 l 0.002)x + (0.024 l 0.006) 0.999 0.09 0.31
HALZ y = (0.173 l 0.002)x + (0.032 l 0.004) 0.999 0.07 0.23
CLZ y = (0.232 l 0.003)x + (0.014 l 0.008) 0.998 0.10 0.34
FNZ y = (0.263 l 0.004)x + (0.029 l 0.014) 0.998 0.15 0.53
LRZ y = (0.229 l 0.003)x + (0.020 l 0.003) 0.999 0.03 0.11
ALZ y = (0.202 l 0.004)x – (0.021 l 0.01) 0.998 0.15 0.50 0.3 – 20
DZP y = (0.328 l 0.003)x + (0.011 l 0.003) 0.999 0.03 0.091 0.2 – 20

Plasma
BRZ y = (0.321 l 0.002)x + (0.000 l 0.02) 0.999 0.08 0.22 0.3 – 20
HTRZ y = (0.207 l 0.002)x + (0.047 l 0.02) 0.999 0.29 0.97
HALZ y = (0.172 l 0.003)x + (0.101 l 0.03) 0.998 0.47 1.57
CLZ y = (0.22 l 0.005)x + (0.136 l 0.015) 0.997 0.21 0.68
FNZ y = (0.262 l 0.004)x + (0.064 l 0.04) 0.998 0.46 1.52
LRZ y = (0.224 l 0.003)x + (0.060 l 0.03) 0.999 0.35 1.16
ALZ y = (0.175 l 0.003)x + (0.039 l 0.03) 0.999 0.43 1.42
DZP y = (0.331 l 0.003)x + (0.071 l 0.03) 0.999 0.27 0.90

Urine
BRZ y = (0.251 l 0.004)x + (0.139 l 0.04) 0.998 0.16 1.59 0.3 – 20
HTRZ y = (0.207 l 0.002)x – (0.014 l 0.02) 0.999 0.29 0.96
HALZ y = (0.183 l 0.002)x – (0.004 l 0.02) 0.999 0.33 1.09
CLZ y = (0.2300 l 0.006)x – (0.002 l 0.006) 1.00 0.08 0.26
FNZ y = (0.264 l 0.001)x – (0.006 l 0.01) 0.999 0.11 0.38
LRZ y = (0.226 l 0.002)x + (0.010 l 0.02) 0.999 0.27 0.88
ALZ y = (0.175 l 0.002)x – (0.005 l 0.03) 0.999 0.34 1.14
DZP y = (0.318 l 0.001)x + (0.011 l 0.01) 0.999 0.09 0.31

Saliva
BRZ y = (0.270 l 0.004)x + (0.040 l 0.02) 0.999 0.11 0.37 0.3 – 20
HTRZ y = (0.200 l 0.001)x + (0.023 l 0.005) 0.999 0.08 0.25
HALZ y = (0.177 l 0.001)x + (0.032 l 0.003) 0.999 0.05 0.17
CLZ y = (0.233 l 0.002)x + (0.031 l 0.009) 0.999 0.12 0.39
FNZ y = (0.254 l 0.004)x + (0.043 l 0.01) 0.999 0.12 0.39
LRZ y = (0.229 l 0.004)x + (0.028 l 0.001) 0.999 0.04 0.13
ALZ y = (0.175 l 0.008)x + (0.049 l 0.002) 0.999 0.14 0.45
DZP y = (0.312 l 0.002)x + (0.042 l 0.009) 0.999 0.09 0.29

3.5.4 Assay precision over a 2 wk period. The concentrations of the analytes in

Precision of a quantitative method is the degree of agree-
ment among individual test results when the procedure
is applied repeatedly to multiple samplings. The analysis
was carried out in one laboratory by one operator, using
the same reagents and instruments over a relatively
short time span. The precision of the method was eval-
uated by assaying QC samples at three different concen-
trations shown in Tables 6 – 9. Intra-assay precision
(within-day repeatability) was evaluated by analyzing the
QC samples (n = 6 for each level) on the same day. The
inter-day precision (between-day reproducibility) is defined
as the long-term variability of the measurement process,
which here was assessed from the same pooled urine,
plasma, and saliva samples as above, by analyzing tripli-
cates of each of the QC samples on six successive days
the quality control samples were calculated using cali-
bration curves. Within- and between-day precision were
assessed by determining the RSDs (% RSD), calculated
from the ratio of SD to the mean of each QC samples of
each concentration level, and expressed as a percentage.
Within- and between-day precision and accuracy data at
the concentrations are shown in Tables 6 – 9. The within-
assay (n = 6) RSD were 0.03 – 4.7%, 1.3 – 7.9%, 2.1 – 6.0%,
2.2 – 5.8% and the between-assay (n = 6) RSD were 0.5 –
7.0%, 3.3 – 7.3%, 2.1 – 7.8%, 2.2 – 8.1% for standard,
plasma, urine, and saliva, respectively.
3.5.5 System suitability
The system suitability was assessed by replicate injec-
tions of the sample at a concentration of 2 ng/lL includ-

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2008, 31, 3704 – 3717 Liquid Chromatography 3713

Table 6. Within- and between-day accuracy and precision data in standards

Analytes Added Within-day (n = 6) Between-day (n = 6)

(ng/lL)
Obtained R (%) RSD(%) Obtained R (%) RSD (%)
(ng/lL) (ng/lL)

BRZ 2.0 2.17 l 0.03 108.6 3.9 1.99 l 0.01 99.7 2.0
4.0 4.05 l 0.04 101.3 3.1 4.01 l 0.05 100.2 4.2
8.0 7.74 l 0.05 96.8 2.3 7.87 l 0.01 98.4 0.5
HTZL 2.0 2.03 l 0.01 101.7 2.7 2.03 l 0.02 101.4 4.5
4.0 4.27 l 0.02 106.8 2.3 4.22 l 0.03 105. 6 3.7
8.0 8.16 l 0.06 102.0 3.6 7.98 l 0.02 99.7 1.3
HAPZ 2.0 2.08 l 0.01 103.9 3.4 2.05 l 0.02 102.4 5.3
4.0 4.31 l 0.02 107.8 2. 3 4.31 l 0.04 107.7 4.8
8.0 8.22 l 0.06 102.8 4.1 7.89 l 0.08 98.6 5.8
CLZ 2.0 2.01 l 0.02 100.5 4.7 2.04 l 0.03 101.9 7.1
4.0 4.37 l 0.02 109.3 1.8 4.25 l 0.05 106.3 5.4
8.0 8.30 l 0.09 103.8 4.6 8.11 l 0.03 101.4 1.6
FNZ 2.0 2.02 l 0.02 100.8 3.7 1.98 l 0.02 99.0 4.2
4.0 4.25 l 0.02 105.9 1.5 4.19 l 0.05 104.7 4.4
8.0 8.04 l 0.09 100.6 3.9 7.91 l 0.05 98.9 2.2
LRZ 2.0 2.02 l 0.01 101.1 2.8 2.02 l 0.03 101.1 5.2
4.0 4.19 l 0.01 104.8 0.1 4.12 l 0.05 102.9 4.7
8.0 7.87 l 0.05 98.4 2.9 7.83 l 0.02 97.9 1.0
APZ 2.0 1.96 l 0.01 98.0 3.6 1.95 l 0.02 97.7 5.8
4.0 4.04 l 0.02 101.1 0.1 3.93 l 0.04 98.2 5.0
8.0 7.71 l 0.02 96.3 1.4 7.74 l 0.04 96.8 2.4
DZP 2.0 2.03 l 0.02 101.3 3.4 2.14 l 0.02 107.2 2.9
4.0 4.30 l 0.02 107.6 1.7 4.18 l 0.06 104.5 4.1
8.0 8.11 l 0.11 101.4 4.0 7.77 l 0.08 97.2 3.2

ing within- and between-day assessments for standard,
spiked plasma, urine, and saliva samples. Precision of
retention times and peak area were examined to evaluate
the system suitability from within-day repeatability
(mean value of six measurements, n = 24) and between-
day precision (mean value of three measurements during
six days, n = 72) expressed by percentage RSD, which
revealed the values of 0.8 – 1.9% and 0.7 – 1.7%, respec-
tively. Precision (RSD) of retention time of the analytes
from chromatograms (n = 6) taken at different buffer
concentrations (0.04 – 0.06 M) was also calculated to eval-
uate the system suitability. The results ranged from 0.5
to 2.2 as shown in Table 10, indicating excellent suitabil-
ity of the system.
3.5.6 Selectivity and specificity
Selectivity of an analytical method is its ability to meas-
ure accurately an analyte in the presence of interference
that may be expected to be present in the sample matri-
ces. The specificity was demonstrated, showing that the
determination of drugs and metabolites was free from
interferences from degradation products. The absence of
any endogenous interfering peak observed in the
extracts of bio-fluids overlapping with any analyte or the
IS indicates the high specificity of the method which can
be used in therapeutic and routine analyses. Typical
chromatograms of blank samples are presented in Figs.
2 – 4.
3.5.7 Robustness
Robustness is the capacity of the method to remain unaf-
fected by small deliberate variations in several chromato-
graphic parameters. The optimum HPLC conditions set
for this method have been slightly modified by the small
changes in flow rate and ammonium acetate concentra-
tion to justify the effect (if any) on the results as a means
to evaluate the method's robustness. It was found that
the percent recoveries of BDZs and metabolites were
excellent under most conditions, and remained unaf-
fected by small deliberate changes of flow rate as shown
in Table 3. Inter-method verification has also been eval-
uated as an evidence of robustness test, when a previ-
ously developed method was applied to measure the
accuracy of six studied parent drugs.
The effect of varying ammonium acetate concentra-
tions from 0.04 to 0.06 M to the peak area, retention
time, resolution factor was not significant and column
performance was not affected.
The study of variation of the flow rate from 1.0 to
1.1 mL/min showed that the peak area decreased at
increased flow rate for all the compounds. This is due to
the fact that at high flow rates the compounds do not
have time to penetrate the pores of particles and to be

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

3714 M. N. Uddin et al. J. Sep. Sci. 2008, 31, 3704 – 3717

Table 7. Within- and between-day accuracy and precision data in plasma samples

Analytes Spiked (ng/lL) Within-day (n = 6) Between-day (n = 6)

Obtained R (%) RSD (%) Obtained R (%) RSD (%)

(ng/lL) (ng/lL)

BRZ 2.0 2.12 l 0.03 105.9 3.9 2.08 l 0.04 103.8 5.3
4.0 3.88 l 0.02 97.1 1.4 3.98 l 0.05 99.6 4.2
8.0 7.87 l 0.09 98.4 3.4 7.91 l 0.15 98.8 5.8
HTZL 2.0 1.99 l 0.01 99.3 1.3 2.14 l 0.02 107.2 4.1
4.0 3.91 l 0.04 97.7 4.6 3.92 l 0.04 97.9 5.3
8.0 7.68 l 0.06 96.0 3.6 7.57 l 0.11 94.6 6.8
HAPZ 2.0 2.08 l 0.04 103.9 7.9 2.11 l 0.02 105.7 3.3
4.0 4.26 l 0.02 106.5 1.9 3.94 l 0.03 98.6 4.2
8.0 7.96 l 0.05 99.5 3.5 7.86 l 0.10 98.3 7.2
CLZ 2.0 2.16 l 0.02 108.2 4.0 2.15 l 0.03 107.6 5.0
4.0 3.99 l 0.8 99.7 7.4 4.02 l 0.05 100.6 4.6
8.0 8.21 l 0.07 102.6 3.4 8.10 l 0.13 101.2 6.7
FNZ 2.0 2.18 l 0.02 109.0 3.6 2.04 l 0.04 102.1 7.1
4.0 4.12 l 0.6 103.0 4.9 4.11 l 0.06 102.9 5.5
8.0 8.02 l 0.07 100.3 3.3 7.78 l 0.15 97.2 7.2
LRZ 2.0 2.03 l 0.01 101.6 2.5 2.06 l 0.04 103.0 6.7
4.0 4.13 l 0.06 107.3 6.3 4.19 l 0.05 104.7 5.4
8.0 8.00 l 0.06 100.0 3.3 7.94 l 0.13 99.3 7.0
APZ 2.0 2.14 l 0.01 107.2 2.5 2.02 l 0.01 101.1 3.4
4.0 4.24 l 0.05 106.1 6.1 4.22 l 0.03 105.6 3.9
8.0 8.24 l 0.04 103.0 2.5 8.19 l 0.11 102.4 7.3
DZP 2.0 2.03 l 0.02 101.4 3.1 2.10 l 0.05 104.9 6.8
4.0 4.02 l 0.06 100.5 4.4 4.01 l 0.07 100.4 4. 9
8.0 7.86 l 0.09 98.2 3.5 7.65 l 0.16 95.6 6.2

retained by the hydrophobic inner surface of the par-
ticles before being eluted by the analytical mobile phase.
Also at high flow rate the retention time of each peak
was reduced with subsequent change in the separation
and resolution factor without affecting their good sep-
aration. However, 1.1 mL/min seems to be a good com-
promise when considering the chromatographic system
and total retention time and precision of peak area.
3.5.8 Stability study
3.5.8.1 Standard solution
The stability of standard solutions was tested by the pro-
posed HPLC method over a period of 90 days. The freshly
prepared solutions at room temperature, and the
90 days stored samples in a refrigerator at 48C, were ana-
lyzed. Chromatograms and recoveries were compared.
No degradation products were present and the drugs are
stable at 48C for at least 90 days, indicating the possibil-
ity of using all studied drugs over a period of 90 days
stored at refrigerator without degradation. Degradation
was decided by using the – 10% criterion.
3.5.8.2 Spiked solution
Freeze – thaw cycle of spiked samples: In case of biologi-
cal samples, plasma, urine, and saliva samples spiked
with 2 ng/lL of analytes were stored deep-frozen at
–258C. Each sample was analyzed once daily for seven
successive days after freeze – thaw cycle for the investiga-
tion of stability when spiked samples were allowed to
thaw at room temperature. Recovery of analytes for the
stored samples was calculated and compared to that of
freshly prepared samples. The results show that up to the
third cycle all analytes were almost stable and up to fifth
cycle, recovery was decreased by less than 10%. Results
have been presented graphically in Fig. 5.
4 Discussion
Although GC methods are more sensitive than HPLC
methods for measuring BDZs and metabolites in biologi-
cal interest, they involve lengthy sample clean-up proce-
dures, and require derivatization steps to increase the
volatility of BDZs, which are thermally unstable under
GC conditions. In addition, they require a suitable IS to
avoid the thermal-molecular rearrangement that occurs
on the column which may lead to poor reproducibility of
the assay.
The proposed chromatographic method has the
advantage to detect and to quantitate the whole spec-
trum of effective compounds and metabolites in a single
run under optimized conditions. HPLC offers the advant-
age to separate and quantitate the compounds without

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2008, 31, 3704 – 3717 Liquid Chromatography 3715

Table 8. Within- and between-day accuracy and precision data in urine samples

Analytes Spiked (ng/lL) Within-day (n = 6) Between-day (n = 6)

Obtained R (%) RSD (%) Obtained R (%) RSD (%)

(ng/lL) (ng/lL)

BRZ 2.0 1.96 l 0.03 98.1 4.4 2.02 l 0.04 101.2 7.4
4.0 3.91 l 0.04 97.8 4.1 4.16 l 0.06 103.9 5.2
8.0 8.57 l 0.08 107.1 3.7 8.46 l 0.13 105.8 5.4
HTZL 2.0 1.96 l 0.01 97.9 2.7 1.91 l 0.02 95.6 5.7
4.0 3.77 l 0.02 94.3 2.7 3.91 l 0.04 97.9 4.6
8.0 7.68 l 0.05 96.0 3.6 7.86 l 0.14 98.3 8.7
HAPZ 2.0 1.95 l 0.01 97.6 2.1 1.93 l 0.02 96.5 4.4
4.0 3.81 l 0.02 95.2 3.6 3.97 l 0.04 99.2 5.5
8.0 7.73 l 0.05 96.6 3.5 7.93 l 0.09 99.1 6.4
CLZ 2.0 1.94 l 0.02 97.2 4.0 1.91 l 0.02 95.6 4.3
4.0 3.78 l 0.03 94.6 3.8 4.06 l 0.04 101.5 4.4
8.0 7.67 l 0.06 95.9 3.7 8.37 l 0.11 104.7 5.7
FNZ 2.0 2.00 l 0.03 100.0 5.8 1.91 l 0.01 95.7 2.1
4.0 3.83 l 0.03 95.9 3.3 3.93 l 0.04 98.4 3.8
8.0 7.63 l 0.12 95.3 6.0 7.97 l 0.13 99.6 6.0
LRZ 2.0 1.97 l 0.01 98.6 3.4 1.89 l 0.01 94.5 2.1
4.0 3.84 l 0.03 95.9 4.1 3.92 l 0.04 98.0 5.0
8.0 7.71 l 0.10 96.4 6.0 8.17 l 0.14 102.1 7.8
APZ 2.0 1.97 l 0.02 98.7 5.2 1.86 l 0.02 93.2 5.1
4.0 3.92 l 0.02 98.1 2.4 4.14 l 0.05 103.5 7.6
8.0 7.72 l 0.05 96.5 3.8 8.32 l 0.12 104.0 8.2
DZP 2.0 1.96 l 0.02 98.2 3.6 1.94 l 0.02 97.3 3.3
4.0 3.94 l 0.05 98.4 3.8 3.91 l 0.06 97.9 5.1
8.0 7.71 l 0.09 96.4 3.9 7.87 l 0.10 98.3 4.1

Figure 5. Stability study data of freeze –

thaw cycles.

derivatization and at low temperature without the risk
of decomposition. It is known that the combination of
HPLC with DAD is considered as a highly effective screen-
ing method. The use of the DAD also gives the advantage
of identifying the analytes both by retention time and
UV spectrum. Furthermore, most of the previously
reported techniques with higher sensitivities required a
complex extraction procedure and the use of large sam-
ple volumes (F 1 mL), which is not practical for pharma-
cokinetic studies, where taking large volumes of blood is
ethically unacceptable.
Usually two hydrolysis procedures, acidic or enzy-
matic, can be used on the conjugates prior to extraction
for urine samples. But BDZs are converted to benzophe-
nones by acid hydrolysis and this may give difficulties to
identify the parent compounds. BDZs are known to bind
the protein albumin but mainly on a-glycoprotein. How-
ever, organic solvents miscible with water, e.g. ACN, can
decrease the solubility of proteins, precipitate them
from aqueous solutions, and render them removable by
centrifugation.

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

3716 M. N. Uddin et al. J. Sep. Sci. 2008, 31, 3704 – 3717

Table 9. Within- and between-day accuracy and precision data in saliva samples

Analytes Spiked (ng/lL) Within-day (n = 6) Between-day (n = 6)

Obtained R (%) RSD (%) Obtained R (%) RSD (%)

(ng/lL) (ng/lL)

BRZ 0.5 0.53 l 0.004 105.3 2.4 0.53 l 0.01 105.4 2.8
1.0 1.01 l 0.01 101.2 4.2 1.06 l 0.03 106.5 8.0
2.0 2.11 l 0.02 105.4 3.8 2.08 l 0.03 104.2 5.6
HTZL 0.5 0.53 l 0.005 107.0 3.8 0.50 l 0.01 100.0 8.1
1.0 1.02 l 0.008 101.6 3.7 0.96 l 0.01 96.4 6.2
2.0 2.00 l 0.01 100.0 2.5 2.15 l 0.01 107.5 2.2
HAPZ 0.5 0.49 l 0.003 97.2 3.0 0.50 l 0.01 100.0 5.5
1.0 0.97 l 0.01 97.0 5.5 0.99 l 0.02 99.0 7.4
2.0 1.98 l 0.008 99.0 2.2 2.14 l 0.01 107.0 2.2
CLZ 0.5 0.51 l 0.005 101.2 3.6 0.49 l 0.01 98.0 6.5
1.0 1.03 l 0.02 103.3 5.7 0.97 l 0.01 97.0 4.9
2.0 2.01 l 0.02 100.5 3.4 2.17 l 0.02 108.6 3.5
FNZ 0.5 0.49 l 0.007 98.8 4.4 0.48 l 0.01 96.0 4.7
1.0 0.99 l 0.02 99.0 5.8 0.96 l 0.01 96.0 3.2
2.0 1.98 l 0.02 98.8 3.1 2.11 l 0.03 105.3 5.4
LRZ 0.5 0.52 l 0.007 103.3 5.3 0.49 l 0.01 98.0 6.4
1.0 1.03 l 0.01 103.0 5.0 1.02 l 0.02 102.4 7.5
2.0 2.04 l 0.02 101.9 3.9 2.09 l 0.02 104.6 3.4
APZ 0.5 0.52 l 0.005 103.1 3.3 0.50 l 0.01 100.0 6.8
1.0 1.01 l 0.01 101.3 5.4 0.98 l 0.01 98.0 6.5
2.0 2.00 l 0.02 100.0 5.5 2.17 l 0.02 108.3 3.7
DZP 0.5 0.49 l 0.008 98.4 4.3 0.49 l 0.01 98.0 7.6
1.0 1.01 l 0.01 101.2 3.2 1.07 l 0.02 106.7 6.1
2.0 1.96 l 0.02 98.0 3.0 2.12 l 0.05 105.8 7.7

Table 10. System suitability and method robustness data

Analytes RSD (%) Buffer effects (CH3COONH4) Inter-method

coefficient
Flow rate Flow rate 0.04 M 0.05 M 0.06 M RSD of of determina-
1.0 mL/min 1.1 mL/min Rt (%) tion

Area Rt Area Rt R (%) Rt R (%) %R

BRZ 4.48 0.43 1.34 1.03 96.8 5.47 l 0.06 99.7 97.7 0.5 0.999
HTZL 3.47 0.56 1.34 1.44 98.6 7.17 l 0.11 101.4 101.4 0.8 –
HAPZ 3.29 0.61 1.36 1.43 99.7 7.66 l 0.11 102.4 104.0 0.7 –
CLZ 4.25 0.39 1.21 1.46 101.5 9.00 l 0.13 99.0 99.2 2.2 0.999
FNZ 3.71 0.56 1.30 1.56 99.4 9.55 l 0.15 99.0 96.3 1.7 1.000
LRZ 1.18 0.63 1.10 1.89 96.7 10.17 l 0.19 101.1 99.4 0.9 0.999
APZ 1.31 0.41 0.70 1.89 97.2 10.77 l 0.21 97.7 98.4 0.8 0.999
DZP 3.98 0.16 1.66 0.79 105.5 15.77 l 0.12 107.2 106.1 1.1 0.999

In SPE method, no interferences were found using the
50:50 mixture of methanol/ACN as eluting solvent to get
a fair chromatogram with expected recoveries of both
BDZs and metabolites when hydrophilic impurities were
completely removed giving clear extract by washing
with 1 mL of water before elution.
Colchicine, having the good resolution and separation
factors between adjacent peaks was used as internal
standard. Freshly prepared quality controlled solutions
were used for analysis because solutions containing IS
after long time storage showed interfering peaks prob-
ably due to the photodecomposition or chemical interac-
tion forming hydrogen bond with the analytes. Alterna-
tively, reconstitution of eluted drugs after dryness on N2
was recommended by the methanolic solution of IS at
certain concentration to avoid the problem.
Analytical methods for BDZs in biological fluids are
problematic, due to the high sensitivity required to
detect them as their low concentration in biofluids. In
addition it is challenging to separate and isolate the
metabolites simultaneously due to their high polarity as
compared to the parent drugs. There are so many active

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2008, 31, 3704 – 3717
metabolites of studied drugs but either they are not com-
mercially available, or at a high cost. Two a-hydroxy
metabolites of alprozolam and triazolam were success-
fully separated and determined.
The method provides satisfactory sensitivity and preci-
sion within considerable analytical range. LOQs for
plasma are supposed to be higher than those in urine
and saliva. This was expected due to the complex chemi-
cal composition from interfering drug release.
5 Conclusions
The HPLC analytical method for the determination of
BDZs in biofluids presented herein meets the criteria for
routine therapeutic drug monitoring or pharmacoki-
netic studies. The advantage of the method over previ-
ously reported methods is the rapidity, simplicity (ease
of sample preparation), high sensitivity, and high selec-
tivity. The LOD and LOQ achieved make the assay appro-
priate for the measurement of therapeutic concentra-
tions.
Saliva should not be seen as a specimen that replaces
the use of other specimens as drugs concentration is too
low. Urine should still be seen as the specimen of choice
for BDZs if evidence of prior exposure to drugs of abuse is
sought (e.g., routine workplace screening). If evidence of
recent use (or abstinence) of drugs is sought then saliva
samples are preferred specimens.
The authors declared no conflict of interest.
6 References
[1] Manual for use by national Laboratories, Recommended Methods
for the Detection and Assay of Barbiturates and Benzodiazepines in Bio-
logical Specimens, ST/NAR/28, United Nations Vienna 1997.
[2] Salem, A. A., Barsoum, B. N., Izake, E. L., Anal. Lett. 2002, 35,
1631 – 1648.
[3] Salem, A. A., Barsoum, B. N., Saad, G. R., Izake, E. L., J. Electroanal.
Chem. 2002, 536, 1 – 9.
[4] Haefely, W., Parnham, M., Bruinvals, J., Discoveries in Pharmacol-
ogy, Vol. 1, Elsevier, Amsterdam 1983, pp. 239 – 306.
[5] Laurence, D. R., Bennett, P. N., Clinical Pharmacology, 5th Edn., Mc
Graw-Hill, New York 1980, pp. 224 – 275.
[6] Linoila, M., in: E. Costa (Ed.), The Benzodiazepines from Molecular
Biology to Clinical Practice, Raven, New York 1983, p. 267.
[7] Gajewska, M., Ciszewska, M. J., Elbieta, M. W., Acta Pol. Pharm.
1984, 41, 213 – 219.
[8] Zimak, J., Volke, J., Gasparic, J., Chem. Listy 1986, 80, 1196 – 1206.
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Liquid Chromatography 3717
[9] Nie, L. H., Liu, D. Z., Yao, S. Z., J. Pharm. Biomed. Anal. 1990, 8, 379 –
383.
[10] Rodriguez-Procopio, J., Hernandez-Hernandez, P., Hernandez-
Hernandez, L., Analyst 1987, 112, 79 – 82.
[11] Duhau, L., Lafarque, P., Levillain, P., Galliot, M., Bourdon, R., Ana-
lusis 1989, 17, 553 – 559.
[12] Nevado, J. J. B., Pe
alvo, G. C., Caldern, M. J. P., J. Chromatogr. B
2002, 773, 151 – 158.
[13] Hernndez, L., Zapardiel, A., Lpez, J. A. P., Bermejo, E., Analyst
1987, 112, 1149 – 1153.
[14] Hemandez, L., Hernandez, P., Blanco, M. H., Analyst 1988, 113,
1719 – 1722.
[15] Van Rooij, H. H., Fakiera, A., Verrijk, R., Anal. Chim. Acta 1985, 170,
153 – 158.
[16] Meatherall, R., Fraser, A. D., Ther. Drug Monit. 1998, 20, 673 – 679.
[17] Dixon, R., J. Pharm. Sci. 1981, 70, 230 – 231.
[18] Ruiz, E., Blanco, M. H., Abad, E., Analyst 1987, 112, 697 – 699.
[19] Klotz, U., J. Chromatogr. 1981, 222, 501 – 506.
[20] Lillsunde, P., Seppala, T., J. Chromatogr. 1990, 533, 97 – 110.
[21] Jiang, Z. L., Tan, J. Y., Yao, L. J., Xing, L. M., Se Pu 2001, 19, 341 –
343.
[22] Cirimele, V., Kintz, P., Mangin, P., Int. J. Legal Med. 1996, 108,
265 – 267.
[23] Borrey, D., Meyer, E., Lambert, W., Van Peteghem, C., De Leenh-
eer, A. P., J. Chromatogr. B Biomed. Sci. Appl. 2001, 765, 187 – 197.
[24] Pirnay, S., Ricordel, I., Libong, D., Bouchonnet, S., J. Chromatogr. A
2002, 954, 235 – 245.
[25] Wilhelm, M., Battista, H. J., Obendorf, D., J. Anal. Toxicol. 2001, 25,
250 – 257.
[26] Uddin, M. N., Samanidou, V. F., Papadoyannis, I. N., J. Liq. Chroma-
togr. Relat. Technol. 2008, 31, 1258 – 1282.
[27] He, W., Parissis, N., J. Pharm. Biomed. Anal. 1997, 16, 707 – 715.
[28] Bogusz, M. J., J. Chromatogr. B 2000, 748, 3 – 19.
[29] Laurito, T. L., Mendes, G. D., Santagada, V., Caliendo, G., de Mor-
aes, M. E. A., De Nucci, G., J. Mass Spectrom. 2004, 39, 168 – 176.
[30] Hegstad, S., Øiestad, E. L., Johansen, U., Christophersen, A. S., J.
Anal. Toxicol. 2006, 30, 31 – 37.
[31] Zweigenbaum, J., Heinig, K., Steinborner, S., Wachs, T., Henion,
J., Anal. Chem. 1999, 71, 2294 – 2300.
[32] Fuh, M., Lin, S., Chen, L., Lin, T., Talanta 2007, 72, 1329 – 1335.
[33] Darius, J., Banditt, P., J. Chromatogr. B 2000, 738, 437 – 441.
[34] Bugey, A., Staub, C., J. Pharm. Biomed. Anal. 2004, 35, 555 – 562.
[35] Luo, Y., Pan, L., Pawliszyn, J., J. Microcolumn. Sep. 1998, 10, 193 –
201.
[36] Pistos, C., Stewart, J. T., J. Pharm. Biomed. Anal. 2003, 33, 1135 –
1142.
[37] Lee, X. P., Kumazawa, T., Sato, J., Shoji, Y., Hasegawa, C., Karibe,
C., Arinobu, T., Seno, H., Sato, K., Anal. Chim. Acta 2003, 492, 223 –
231.
[38] Uddin, M. N., Samanidou, V. F., Papadoyannis, I. N., J. Sep. Sci.
2008, 31, 2358 – 2370.
[39] Samyn, N., De Boeck, G., Cirimele, V., Verstraete, A., Kintz, P., J.
Anal. Toxicol. 2002, 26, 211 – 215.
[40] Miller, J. C., Miller, J. N. (Eds.), Statistics for Analytical Chemistry,
Vol. IV, Horwood E, Chichester 1986.
www.jss-journal.com