Full text of DOI:10.1093/chromsci/41.6.301

Journal of Chromatographic Science, Vol. 41, July 2003
Determination of Trace RA by Capillary Electrophoresis–
Solid-Phase Microextraction with Direct UV Detection
Mei Fang* and Guowang Xu
National Chromatographic R&A Center of China, Dalian Institute of Chemical Physics, the Chinese Academy of Sciences, Dalian,
116011, P.R. China
ment (preconcentration) role, can lower the concentration
detection limit (7–10). The preparation of setup for offline
Abstract
A new method of offline solid-phase microextraction coupled with
capillary electrophoresis (CE) [in an uncoated capillary using borax
and boric acid in acetonitrile–water (40:60) as the running buffer]
with UV detection at 254 nm is presented and used for direct
determination of trace biomedically important retinoic acid (RA)
without any derivation. As low as 138 ng/mL RA in aqueous
solution can be determined using CE with UV detection with a
concentration factor of 19.6 times.
SPE is easier and simpler than that for online SPE, but only a
few methods have been reported that describe offline SPE for
retinoids (1). In addition, the coated capillary may be used to
suppress electro-osmatic flow, improving the separation per-
formance (11), but commercially coated capillaries are expen-
sive and have very short lives (12). On the other hand, a bare
(uncoated) capillary also has some advantages (12–15).
The present paper is a new report concerning offline solid-
phase microextraction (SPME) coupled to bare CE with
native UV detection for determination of trace RA without
derivitization.
Introduction
Retinoic acid (RA) and its metabolites have an important role
in nutrition, vision, and in the treatment of various diseases in
oncology and dermatology. On the other hand, it is also a toxic
compound that causes mild hypervitaminosis A and, even
more, tetratogeny. Therefore, the determination of RA is very
necessary. RA is a monocyclic parent compound containing five
carbon–carbon double bonds and a functional terminal carboxy
group at the terminus of the acyclic portion, and thus it has
favorable UV characteristics (1). It is sensitive to light, heat, and
oxygen and is insoluble in aqueous solutions. Most commercial
capillary electrophoresis (CE) instruments are equipped with
UV–vis detectors, but the detection of native absorbance is not
sensitive enough (2) to trace RA in very low sample volumes.
The concentration detection limits of UV absorbance in CE
often fall in the µmol/L range (3). Consequently, detection
sensitivity has become the “bottle-neck” in CE analysis of trace
components in small-volume situations (4). The capillary-
modified Z-cell gives only a sixfold signal-to-noise ratio because
of light attenuation, despite a 60-times longer path length (5).
A double-stacking procedure based on field enhancement is
described as a means to increase the concentration sensitivity
in CE (6). Solid-phase extraction (SPE), because of its enrich-
Experimental
Instruments
The CE instrument used was an NT-binda 1229-HPCE ana-
lyzer made by the Beijing Institute of New Technology Appli-
cation (Beijing, China).
Reagents
The RA solution was made by dissolving 1.400 mg RA (bio-
chemical reagent) with 95% ethanol to obtain a 690-µg/mL
stock solution. The solution was diluted to 13.8 µg/mL with
methanol. Then, 1380 and 138 ng/mL solution were prepared
by dilution with water. Borax–boric acid running buffer was
prepared by dissolving 0.380 g borax and 0.310 g boric acid in
100 mL acetonitrile–water (40:60) solution.
Construction of SPME setup
A 1.5-cm long polytetrafluoroethylene (PTFE) tube with a
300-µm i.d. was cut. The inside diameter of the tube was
expanded by inserting an iron thread of o.d. 360 µm. A 0.5-mm
thick glass fiber was inserted, followed by a 5-cm long capillary
with a 75-µm i.d. From the other end of the PTFE tube, some
40-µm Accubond ODS solid-phase packing (Agilent Technolo-
gies, Palo Alto, CA) was filled dryly and compacted up to 5-mm
*Author to whom correspondence should be addressed. Current address: Pharmacy College of
West China, Sichuan University, c/o Professor Guozhen Fang, P.O. Box 55, College of
Chemistry,Sichuan University, No. 29 Wangjiang Road, 610064, P.R. China.
Email mfangcd@yahoo.com.
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.
301

Journal of Chromatographic Science, Vol. 41, July 2003
thickness. A 0.5-mm thick glass fiber and 5-cm long capillary
with 75-µm i.d. were then inserted one after another.
Detection of 1380 ng/mL RA solution. A comparison of the
electropherograms of 1380 ng/mL RA solution with and
without using the combination of CE and SPME setup (by pre-
concentrating for 11 min) are shown in Figure 2. It can be seen
that only if the stated combination was adopted, as low as 1380
ng/mL of RA can be detected. Otherwise, the aim could not be
achieved, unless other effective methods were taken.
CE–SPME procedure
The electrophoresis of RA standard solution was performed
in borax–boric acid running buffer. The capillary used was 36-
cm long with 375-µm o.d. and 75-µm i.d. The running voltage
was 15 kV. Preconcentrated RA solution with SPME made by
the authors was injected for 40 s by a 6.5-cm elevation of the
capillary. The detection wavelength was 254 nm.
Detection of 138 ng/mL RA solution. A lower concentration
of RA was tested. The method was the same as the previous,
except that preconcentration was stopped after 60 min (to
increase a further level of preconcentration). Figure 3 shows
that as low as 138 ng/mL RA solution was determinable under
the experimental conditions, whereas without the SPME com-
bination, it was impossible.
Results
Ordinary CE of RA standard solution
The electrophoresis of 13.8 µg/mL RA standard solution was
borax–boric acid running buffer. The conditions were a run-
ning voltage of 15 kV and anodic hydrodynamic injection for
40 s by 6.5-cm elevation of the capillary. The electropherogram
detected at 254 nm is shown in Figure1A. As low as 2700
ng/mL of RA cannot be detected (Figure 1B).
Discussion
Precision of RA detection
RA (13.8 µg/mL) was electrophoresized five times repeatedly
by CE–UV without SPME. The peak time of RA ranged from
3.3 to 3.7 min. The precision of detection of RA was good.
Highly sensitive detection of RA
A combination of CE with an SPME setup was developed to
detect the more trace RA. The procedure was as follows.
Pretreatment of SPME
After connecting a syringe holding solution with the SPME
setup by means of a PTFE tube as the connector, 10 drops
each of acetonitrile, methanol, and water were pushed out
from the SPME setup column in order to wash and pre-equi-
librate it.
Sampling
RA solution (1380 ng/mL) was aspired into a 5.00-mL scale
injector. The solution was pushed through the SPME auto-
matically for 11 min by means of pressure equipment. The
SPME setup was taken off the syringe, one section of the cap-
illary in it was pulled out, and the SPME setup column was
back washed with water. After putting back the sec-
tion of capillary to the SPME setup, the water in it was
pushed dry with vacant syringe.
Figure 1. Electropherograms of 13.8 µg/mL (A) and 2700 ng/mL (B) of RA
standard solutions by CE–UV without SPME.
Elution
Methanol was aspired into the syringe and con-
nected with the PTFE tube as the connector. After
adjusting the liquid level in the syringe, the syringe
with the PTFE tube connector was connected with
the SPME column. By observing the scale on the
syringe, a quantity of 2 µL methanol was pushed for-
ward through the SPME setup, and the eluate was col-
lected in a microvial.
CE detection
The mentioned methanol elution solution was elec-
tropherosized according to the previously described
method.
Figure 2. Comparison of electropherograms of 1380 ng/mL of RA solution obtained by
CE–UV with (A) and without (B) combination of CE and SPME.
302

Journal of Chromatographic Science, Vol. 41, July 2003
Effect of elution volume on detection
RA (1380 ng/mL) was electrophoresized and detected after
having been absorbed by SPME and eluted. The effect of the
eluate volume on detection was investigated by collecting cer-
tain volumes of eluates. In the case of Figure 4, 1.40, 1.50, and
1.50 µL (curves A, B, and C, respectively) of RA eluates were
collected one after another and electrophoresized. From Figure
4, it may be seen that first eluate (1.4 µL) already contained all
of the separated RA.
It was anticipated that the first collection of 2.0 µL of eluate
would be more suitable for detecting RA with less reading
error of the volume and higher repeatability than the 1.4-µL
eluate, and with enough sensitivity. Another 1380 ng/mL of
standard solution was sampled for 15 min, and two 2.0 µL of RA
eluates were sequentially collected and then electrophoresized.
The experimental results shown in Figure 5 confirm that the
first 2.0-µL eluate volume was suitable (it contained all RA
separated).
Figure 3. Detection of 138 ng/mL RA solutions by CE–UV after pushing
them through the SPME for 60 min (A) and without SPME (B).
Reproducibility of detecting low concentrations
(1380 ng/mL) of RA
After a 1380-ng/mL RA standard solution was sampled and
eluted from the SPME column by methanol (4.0 µL in all), for
which the electropherogram is shown in Figure 5, another
1380-ng/mL RA solution was directly pushed through the pre-
viously described SPME column at once, without any pre-equi-
libration with water. The resulting electropherogram is shown
in Figure 6. Comparing Figures 5 and 6, the detection of the
low concentration (1380 ng/mL) of RA was confirmed with
good reproducibility using this SPME column.
Effect of methanol elution on the CE detection of RA
Usually the organic solvent (i.e., methanol) contained in a
sample can improve the detection sensitivity of the sample
because of the stacking effect of CE. However, in the case of
Figure 7, a comparison of the electropherograms of 1380
ng/mL RA diluted with water and detected by CE–UV (curve
A), diluted with methanol and detected by CE–UV (curve B),
and diluted with water and detected by CE–SPME–UV (curve
C) was made. It can be seen from Figure 7 that combining
CE with SPME was key for im-
Figure 4. The effect of eluate volumes of RA standard on electropherograms
obtained by CE–UV–SPME (A, B, and C represent electropherograms of
first, second, and third consecutive eluates, respectively).
proving the detection sensitivity of
common UV spectrophotometry,
whereas the addition of methanol
alone and without the cooperation
of SPE could not, unless another
effective supplementary method was
taken.
Test of blank solution
A blank solution without RA was
tested by the CE–SPME–UV method.
The experimental results showed that
there was no peak in the electro-
pherogram obtained from the water
blank. The result proved that there
was no interference from the packing,
tube, etc. for the determination of RA
in samples.
Figure 5. The effect of eluate volumes of RA standard on electropherograms obtained by CE–UV–SPME
(A and B represent eluates of the first 2.0 µL and consecutive second 2.0 µL, respectively).
303

Journal of Chromatographic Science, Vol. 41, July 2003
Conclusion
Acknowledgment
Compared with online SPME, the offline SPME is relatively
easy and convenient to make, and its effect is also satisfactory
in the preconcentration of trace components. Specifically, the
sampling is unimpeded because the offline SPME column is
not blocked from troublesome bubbles and high pressure
caused by sampling into the very thin and long capillary. Fur-
thermore, the combination of uncoated CE and offline SPME
with ordinary UV detection carries respective advantages and
may make up each other’s deficiencies. The combination pro-
cedure has been applied to enrich and measure native
absorbance of trace (as low as 138 ng/mL) RA without any
chemical derivation. The whole operation requires less than 1
h every time, and this one column may be used 3 times.
Because of the facility of make-and-adjustable-length charac-
teristics of the SPME setup column, the method is practical
and robust.
This work was supported by a grant from the National
Science Foundation of China.
References
1. R. Wyss. Chromatographic and electrophoretic analysis of bio-
medically important retinoids. J.Chromatogr. B 671: 381–425
(1995).
2. K. Swinney and D. Bornhop. A review of CE detection method-
ologies. Crit. Rev. Anal. Chem. 30: 1–30 (2000).
3. D.M. Pinto, E.A. Arriaga, D. Craig, J. Angelova, N. Sharma,
H. Ahmadzadeh, and N.J. Dovichi. Picomolar assay of native pro-
teins by capilllary eletrophoresis precolumn labeling, submicellar
separation, and laser-induced fluorescence detection. Anal.
Chem. 69: 3015–21 (1997).
4. S. Nie, R. Dadoo, and R.N. Zare. Ultrasensitive fluorescence
detection of polycyclic aromatic hydrocarbons in capillary elec-
trophoresis. Anal. Chem. 65: 3571–75 (1993).
In short, this is an example of a method using offline SPME
to enhance the concentration sensitivity in CE, and it is rec-
ommended.
5. P. Jandik and W.R.Jones. Optimization of detection sensitivity in
the capillary electrophoresis of inorganic anions. J. Chromatogr.
546: 431–43 (1991).
6. S. Palmarsdottir and L.-E. Edholm. Enhancement of selectivity
and concentration sensitivity in capillary zone electrophoresis by
on line coupling with column liquid chromatography and uti-
lizing a double stacking procedure allowing for microliter injec-
tions. J. Chromatogr. A 693: 131–43 (1995).
7. M.A. Strausbouch, S.J. Xu, J.E. Ferguson, M.E. Nunez, D.
Machacek, G.M. Lausan, P.J. Wettstein, and J.P. Landers. Con-
centration and separation of hypoglyemic drugs using solid-phase
extraction-capillary eletrophoresis. J. Chromatogr. A 717: 279–91
(1995).
8. E. Bonneil and K.C. Waldron. Characterization of a solid-phase
extraction device for discontinuous online preconcentration in
capillary electrophoresis-based peptide mapping. J. Chromatogr.
B 736: 273–87(1999).
9. J.R. Veraart, C. Gooijer, H. Lingeman, N.H. Velthorst, and U.A.T.
Brinkman. At-line solid-phase extraction coupled to capillary
eletrophoresis: determination of amphoteric compounds in bio-
logical samples. J. High Resol. Chromatogr. 22: 183–87 (1999).
10. H. Soini, T. Tsuda, and M. Novotny. Electrochromatographic
solid-phase extraction for determination of cimetidine in serum by
micellary electrokinetic capillary chromatography. J. Chromatog.
559: 547–58 (1991).
Figure 6. Electropherogram detected by the combination of CE–SPME
setup. The electropherogram of 2 µL elution solution of 1380 ng/mL RA
without pre-equilibration (A) and that of the methanol elution as a starting
contrast (B).
11. Mei Fang, Rui Zhao, Li Shen, Tianshen Su, and Guoquan Liu. Sep-
aration of four basic proteins in the mixture by capillary elec-
trophoresis with a new chemical modification column. Anal.
Lett. 35: 397–411 (2001).
12. Z.K. Shihabi. Capillary electrophoresis of double-stranded DNA
in an untreated capillary. J. Chromatogr. A 853: 349–54 (1999).
13. E.N. Fung and E.S. Yeung. High-speed DNA sequencing by using
mixed poly(ethylene oxide) solutions in uncoated capillary
columns. Anal. Chem. 67: 1913–19 (1995).
14. H.M. Liebich, G. Xu, C. Di Stefano, R. Lehmann, H.U. Haring, P.
Lu, and Y. Zhang. Analysis of normal and modified nucleosides
in urine by capillary electrophoresis. Chromatographia 45:
396–401 (1997).
15. O.D. Camejane, J.J. Schwinefus, S.-C. Wang, and M.D. Morris.
Electrophoretic separation of linear and supercoiled DNA in
uncoated capillary. J. Chromatogr. A 849: 267–76 (1999).
Figure 7. Comparison of electropherograms of 1380 ng/mL RA solution
obtained under different conditions. A, B, and C represent RA diluted
with water and detected by CE–UV, RA diluted with methanol and detected
by CE–UV, and RA diluted with water and detected by CE–SPME–UV,
respectively.
Manuscript accepted May 23, 2003.
304