ORIGINAL PAPER

Development of a novel luminol chemiluminescent methodcatalyzed by gold nanoparticles for determinationof estrogens

Yongxin Li & Ping Yang & Po Wang & Lun Wang

Received: 4 August 2006 /Revised: 3 October 2006 /Accepted: 10 October 2006 / Published online: 22 November 2006# Springer-Verlag 2006

Abstract It has been found that gold nanoparticles (nano-Au) enhance the chemiluminescence (CL) of the luminol–hydrogen peroxide system and that estrogens inhibit theseCL signals in alkaline solution. CL spectra, UV–visiblespectra, X-ray photoelectron spectra (XPS), and transmis-sion electron microscopy (TEM) were used to investigatethe mechanism of the CL enhancement. On the basis of theinhibition, a flow-injection CL method has been establishedfor determination of three natural estrogens. Under theoptimized conditions, the linear range for determination ofthe estrogens was 0.07 to 7.0 μmol L−1 for estrone, 0.04 to10 μmol L−1 for estradiol, and 0.1 to 10 μmol L−1 forestriol. The detection limits were 3.2 nmol L−1 for estrone,7.7 nmol L−1 for estradiol, and 49 nmol L−1 for estriol, withRSD of 2.9, 2.6, and 1.8%, respectively. This method hasbeen used for analysis of estrogens in commercial tabletsand in urine samples from pregnant women.

Keywords Gold nanoparticles . Luminol . Estrogens .

Chemiluminescence

Introduction

Light emission induced by chemical reactions, known aschemiluminescence (CL), has been intensively investigatedfor many years. Traditionally, the study of CL was limitedto molecular systems [1, 2]. Recently, the increasingavailability of nanoparticles has attracted widespreadattention in CL analysis, because of their high surfaceareas, high activity, and high selectivity. For example, Bardet al. found that semiconductor nanoparticles could emitlight in chemical and electrochemical reactions [3–5]. Itwas proposed that the special electronic structures andsurface properties of the semiconductor nanoparticles werethe origin of the emission [3–5]. Pavlov and co-workershave reported biocatalytic generation of chemiluminescenceinvolving gold nanoparticles in the presence of luminol andhydrogen peroxide, and its use for amplified detection ofDNA or telomerase activity [6]. Cui and co-workers havealso found that gold nanoparticles of different sizes couldact as catalysts to enhance the luminol–hydrogen peroxideand bis(2,4,6-trichlorophenyl) oxalate –hydrogen peroxideCL reactions [7, 8].

Gold nanoparticles (nano-Au) have been widely used inrecent decades; their catalysis of gas phase [9–11] andliquid-phase [12] redox reactions is a new area of research.According to Lopez’s report [13], there might be severaleffects contributing to the special catalytic properties ofsupported nanosized gold particles, and the most importanteffect was related to the availability of many low-coordinated gold atoms on the small particles. Interactionbetween oxide materials and nano-Au was, therefore, notessential for the catalytic activity of the nano-Au [14], and

Anal Bioanal Chem (2007) 387:585–592DOI 10.1007/s00216-006-0925-0

Y. Li (*) : P. Yang : P. Wang : L. Wang (*)Anhui Key Laboratory of Functional Molecular Solids,College of Chemistry and Materials Science,Anhui Normal University,Wuhu, 241000 Anhui, Chinae-mail: liyx72@yahoo.com

L. Wange-mail: wanglun@mail.ahnu.edu.cn

some unsupported gold nanoparticles, for example goldcolloids might be directly used as catalysts for some CLreactions [7, 8, 15] in liquid-phase systems. Although goldcolloids have been used in some liquid-phase CL systems,real applications of this novel catalytic CL phenomenon arefinite [16] and it is necessary to explore some newapplications of gold colloids in liquid-phase CL systems.

Estrogens have been widely used to regulate women’sdiseases in medicine. Determination of estrogens inbiological matrices usually requires analytical method witha low detection limit, because the concentrations of estro-gens in real samples is usually low. Many analyticalmethods have been used for determination of estrogens,for example enzyme-linked immunoadsorbent assay(ELISA) [17], chromatographic methods [18, 19], electro-chemical methods [20, 21], and fluorimetry [22]. Althoughmost of these methods are specific and sensitive, they arealso expensive and complicated. With the advantages ofhigh sensitivity, rapidity, wide linear working ranges, andrelative inexpensive apparatus and reagents, chemilumines-cence (CL) methods have also been used to detect estro-gens. In 1986, Van den Berg et al. [23] developed a CLimmunochemical method after HPLC for detection ofestrogenic hormones. Lin et al. [24] recently established amicro-plate magnetic CL enzyme immunoassay for rapidand high-throughput analysis of estradiol in water samples.Xie et al. [25] developed a method in which acidicpotassium permanganate–formaldehyde CL was used formeasurement of diethylstilbestrol. Wang et al. [26] reporteda tetrasulfonated cobalt phthalocyanine (CoTSPc) catalyzedluminol–H2O2 CL system for the determination of diethyl-stilbestrol. Although many methods have been establishedfor the determination of estrogens, the complicated proce-dures and poor selectivity still proscribe application of thesemethods to analysis of biological samples. It is, therefore,necessary to develop a novel method for determination ofestrogens in real samples.

In this work, a colloidal solution of gold nanoparticles(nano-Au) were synthesized by the citrate reduction method[27] and used to catalyze the luminol–H2O2 CL reaction.Estrogens greatly inhibit this CL emission in alkalinesolution, with the extent of the inhibition depending onthe concentration of estrogens. This inhibition was there-fore combined with a flow-injection system to establish anovel, highly sensitive CL method for determination ofestrogens. The possible CL reaction mechanism is dis-cussed in detail. In comparison with our previous work[28], nano-Au was easier to synthesize; it is the mostwidely used tag in biotechnological systems because of itsexcellent physical and chemical characteristics [29]. Itcould therefore be used as a biological label to developnovel CL immunoassay methods with extensive potentialapplications in analytical and bioanalytical chemistry.

Experimental

Reagents and solutions

A 10 mmol L−1 stock solution of luminol (3-amino-phthalhydrazide) was prepared by dissolving 443 mgluminol (Sigma) in 250 mL 0.1 mol L−1 NaOH solutionwithout purification. Working solutions of luminol(0.1 mmol L−1, 100 mL) were prepared by diluting thestock solution. working solutions of H2O2 (30 mmol L−1,100 mL) were prepared fresh daily from 30% (w/w) H2O2

(Shanghai Reagent Company, Shanghai, China).HAuCl4.4H2O was purchased from Shanghai Reagent

Company. A 20 mmol L−1 HAuCl4 stock solution wasprepared by dissolving 820 mg HAuCl4.4H2O in water in a100-mL calibrated tube. Trisodium citrate stock solution(38.8 mmol L−1) was prepared by dissolving solidtrisodium citrate (Sanpu Chemical, Shanghai, China) indouble-distilled water. All the reagents were of analyticalgrade, and double-distilled water was used throughout.

Synthesis of gold nanoparticles

Gold nanoparticles were prepared by the Frens method[27], with slight modification. Briefly, HAuCl4 solution(0.01%, 100 mL) was boiled and trisodium citrate solution(1.0%, 2.5 mL) was added quickly with vigorous stirring.The color of the solution changed from yellow to red in afew seconds. The solution was heated under reflux for15 min, cooled naturally to room temperature, filteredthrough a 0.45-μm Nylon membrane, diluted to 100 mL,and the pH was adjusted to 9.0 with K2CO3. The averagediameter of the particles was approximately 40 nm (Fig. 1)and the absorbance maximum appeared at 538 nm.

Fig. 1 TEM of gold nanoparticles

586 Anal Bioanal Chem (2007) 387:585–592

Apparatus

All CL measurements were performed with the modelIFFM-E flow-injection CL analysis system (Xi’an Remax,Xi’an, China). CL spectra were measured on a model F-4500 spectrofluorimeter (Hitachi, Tokyo, Japan). UV–visible spectra were acquired with a model UV-3010spectrophotometer (Shimadzu, Japan). X-ray photoelectronspectroscopy (XPS) was performed with an Escalab MKIIspectrometer (VG, UK).

Procedures

A schematic diagram of the flow-injection chemilumines-cence system is shown in Fig. 2. One peristaltic pump (twochannels) was used for the gold nanoparticles or formixtures of the gold nanoparticles and estrogen; anotherpump was used to carry luminol solution and hydrogenperoxide solution. PTFE tubing (0.8 mm i.d.) was used toconnect all components in the flow system. Injection wasperformed using a six-way injection valve fitted with a 150-μL sample loop. The CL signal was recorded and the datawere processed automatically by Remax software underWindows 98.

Gold nanoparticles solution or a mixture of estrogen andgold nanoparticles solution was injected into the aqueouscarrier stream, by use of the six-way valve, then mixed withluminol and hydrogen peroxide. The mixed solution wastransferred to the CL cell and the CL signal was thenrecorded automatically by use of Remax software. Therelative CL intensity ΔI (ΔI = I0 − I, where I0 stands for thesignal in the absence of estrogens and I stands for the signalin the presence of estrogens) showed the effect of estrogenon the CL intensity of the luminol–H2O2–gold nano-particles system; this was used for quantitative analysis ofthe estrogens.

CL spectra were measured using a model F-4500spectrofluorimeter combined with a flow-injection system,the light-entrance slot of which was shut.

Application to urine samples

Urine samples (supplied by Yijishan Hospital, Wuhu,China.) obtained from normal pregnant women were frozenand stored until analysis. Before CL analysis urine samples(4 mL) were centrifuged in ultrafiltration centrifuge tube at4 °C and 10,000 rpm for 30 min. Urine (15 μL) was mixedwith 0.5 mL gold nanoparticles solution in a 10-mLcalibrated tube, and the mixture was diluted to volumewith double-distilled water and mixed thoroughly. Subse-quent steps were similar to those described above.

Results and discussion

Inhibition of the luminol CL reaction by estrogens

The luminol–hydrogen peroxide CL system is a popularCL system which emits weak CL in alkaline solution in theabsence of a catalyst (Fig. 3c). When 50 μmol L−1 goldnanoparticles were added into this CL system, however, theCL signal was enhanced greatly (Fig. 3a). It can thereforebe assumed that gold nanoparticles strongly catalyze theluminol–H2O2 CL reaction. When 1.0 μmol L−1 estriol wasmixed with the gold nanoparticles and injected into thestream, however, the CL intensity decreased dramatically(Fig. 3b), and similar phenomena were also observed forestradiol and estrone. Because the relative CL intensity(ΔI) was proportional to the concentration of estriol, thephenomenon could be used as an inhibitive CL method forassay of some estrogens.

PMT COM

ab

cd

P

V

F

W

HVM

Fig. 2 Schematic diagram of the flow-injection system. a, goldnanoparticles or mixture of gold nanoparticles and estrogen (sample)solution; b, H2O; c, luminol solution; d, H2O2 solution; P, peristalticpump; V, six-way injection valve; M, mixing coil; W, waste; F, flow-cell; PMT, photomultiplier tube; HV, negative voltage; COM,computer

0 5 10 15 20 250

100

200

300

400

c

b

a

Rel

ativ

e C

L In

ten

sity

time (s)Fig. 3 Time course of the kinetic profiles of the luminol–H2O2 CLreaction, without gold nanoparticles (c) and with the particles in theabsence of estriol (a) and in the presence of 1.0 μmol L−1 estriol (b).Conditions: luminol, 0.1 mmol L−1; H2O2, 30 mmol L−1; NaOH,50 mmol L−1; gold nanoparticles, 50 μmol L−1; flow rate, 2.8 mLmin−1

Anal Bioanal Chem (2007) 387:585–592 587

Studies of the emitting species

The CL mechanism of the luminol–H2O2 reaction has beendiscussed by Merényi and Burdo [30, 31]. They confirmedthat the excited sate of 3-aminophthalate was the emitter inthe luminol–H2O2 CL reaction system and that themaximum emission of the CL reaction was at 425 nm[32]. During oxidation of luminol oxygen-related radicals(for example, OH·, O2��, and others), as oxidants, areexpected to be generated from H2O2. To confirm theidentities of the emitting species in this estrogen inhibitingCL reaction catalyzed by gold nanoparticles, CL spectrawere measured during the process. The CL spectra obtainedare shown in Fig. 4. The CL spectra of luminol–H2O2

(Fig. 4c) and luminol–H2O2–gold nanoparticles systems inthe presence (Fig. 4b) and absence (Fig. 4a) of estrogenwere examined. It is apparent the CL maximum is locatedat approximately 425 nm and that the CL intensitydecreased when estriol was added, in agreement with theemission maximum of the luminol–hydrogen CL reaction(Fig. 4a).

XPS and UV–visible absorption spectroscopy were usedto obtain more information about the emitting species. XPSresults, shown in Fig. 5, revealed the binding energy ofAu4f for gold nanoparticles before and after the CLreaction were similar, indicating that the oxidation state ofgold was not involved in catalysis of the CL process; theorigin of the CL was therefore assigned to gold nano-particles themselves. According to Zhang [8], when goldnanoparticles are added to the luminol–H2O2 CL system inalkaline solution, H2O2 might be absorbed on the surface ofthe gold nanoparticles, causing partial electron transferfrom the gold nanoparticles to the adsorbed H2O2, and the

O–O bond of H2O2 might be broken into two OH. radicalswhich react with luminol and HO�

2 to facilitate formation ofluminol radicals and superoxide radical anion O2��ð Þ [33].Further electron-transfer processes between luminol radi-cals and O2�� radicals would produce hydroxyhydroper-oxide [34] on the surface of the gold nanoparticles, leadingto enhancement of CL intensity.

UV–visible absorption spectra of estriol, gold nano-particles, and their mixture are shown in Fig. 6. It can beseen that gold nanoparticles have a notable absorption peakat approximately 538 nm and that estriol had two distinctabsorption peaks at approximately 235 and 295 nm. Whenestriol was mixed with gold nanoparticles no new absorp-tion peaks appeared, and the absorption spectrum of themixed system was approximately the sum of two absorptionspectra except that the absorption intensity of the gold

350 400 450 500 550

0

20

40

60

80

100

c

b

a

Rel

ativ

e C

L In

tens

ity

Wavelength /nmFig. 4 Chemiluminescence spectra of the luminol–hydrogen peroxidesystem. (a) luminol + H2O2 + nano-Au; (b) luminol + H2O2 + nano-Au + estriol; (c) luminol + H2O2. Conditions: luminol, 0.1 mmol L−1;H2O2, 30 mmol L−1; NaOH, 50 mmol L−1; gold nanoparticles,50 μmol L−1; estriol, 1.0 μmol L−1; flow rate, 2.8 mL min−1

200 300 400 500 600

0.0

0.5

1.0

c

b

aAb

sorb

ance

/ a.

u.

Wavelength / nmFig. 6 UV–visible absorption spectra of: (a) gold nanoparticles; (b)estriol; (c) a + b. Conditions: gold nanoparticles, 50 μmol L−1; estriol,1.0 μmol L−1

80 85 90 95

Rel

ativ

e In

tens

ity /

a.u.

Binding Energy / eV

gold before reactiongold after reaction

Fig. 5 XPS of gold nanoparticles before and after the CL reaction

588 Anal Bioanal Chem (2007) 387:585–592

nanoparticles absorption peak decreased at approximately538 nm. It is therefore apparent that no complex wasformed between gold nanoparticles and estriol, but thatestriol interact in some way with the gold nanoparticles.Ghosh and Vijay Sarathy [35, 36] have also reported thatcompounds containing alcohol groups could interact weak-ly with gold particles. Taking into consideration adsorptionof H2O2 on the surface of gold nanoparticles, we can

deduce there must be competitive adsorption on the surfaceof gold nanoparticles between estrogens and H2O2, whichleads to inhibition of CL. Estrogens can, on the other hand,react with hydrogen peroxide or oxygen-related radicalsgenerated from H2O2, which might consume the oxidantand inhibit CL intensity. The possible CL reactionmechanism can therefore be summarized by the equationsin Scheme 1.

O

NH

NH

O NH2

+ OH

O

N

O

NH

NH2

+ H 2O_

H2O2+ OH HO2 + H2O_ _

-

AuO

O

H

H

AuO

O

H

H

..

HO2-

NNH

O

O

NH2

-

Au

O2.-

NN

O

O

NH2

-.

N

N

O2

O

O

NH2NH2

COO-

COO-N2

NH2

COO-

COO-NH2

COO-

COO-

Au +HO

R

R'

R

R'

AuO

H

Estrone: R', =O; R, H

Estradiol: R', -OH; R, H

- -*

+

*

+ light (425 nm)

Estriol: R', -OH; R, -OHScheme 1 Summary of possible CL reaction mechanism

Anal Bioanal Chem (2007) 387:585–592 589

Optimization of the experimental conditions

Effect of luminol concentration

The concentration of luminol in the reagent solution is animportant factor affecting signal intensity. To investigatethe effect of luminol concentration on CL emissionintensity a study was performed in the range 2.5 μmolL−1 to 2.5 mmol L−1 luminol for the three natural estrogens.Results (Fig. 7) showed that the greatest CL emissionintensity was obtained with 0.1 mmol L−1 luminol. Theconcentration of the luminol solution was therefore adjustedto 0.1 mmol L−1.

Effect of H2O2 concentration

The effect of hydrogen peroxide concentration on relativeCL intensity was investigated in the range 5.0–107 mmolL−1. Figure 8 shows that for both estradiol and estriol thegreatest CL intensity was obtained with 27 mmol L−1 H2O2

and that for estrone the greatest CL intensity was obtainedwith 80 mmol L−1 H2O2; 30 mmol L−1 H2O2 was used insubsequent experiments.

Effect of NaOH concentration

Because of the nature of luminol reaction, which is morefavored under basic conditions, an alkaline medium wasused to improve the sensitivity of the system. Figure 9shows that the CL intensity increased with increasingNaOH concentration but that reproducibility was also worseat concentrations above 0.1 mol L−1. The maximum

response was obtained for 50 mmol L−1 NaOH; this was,therefore, chosen as optimum.

Effect of flow rate

To obtain satisfactory emission intensity, the effect of theflow rate of reagent solution was examined. The CLemission intensity was found to increase constantly withincreasing flow rate, because of the rapid rate of the CL

0 2 4 6 8 1060

90

120

150

180

c

b

a

Rel

ativ

e C

L In

tens

ity

Cluminol

(× 10 - 4 mol/l)Fig. 7 Effect of luminol concentration on CL intensity. (a) luminol +H2O2 + gold nanoparticles + estrone; (b) luminol + H2O2 + goldnanoparticles + estradiol; (c) luminol + H2O2 + gold nanoparticles +estriol. Conditions: H2O2, 30 mmol L−1; NaOH, 50 mmol L−1; goldnanoparticles, 50 μmol L−1; estriol, 1.0 μmol L−1; flow rate, 2.8 mLmin−1; estrone, 1.0 μmol L−1; estriol, 1.0 μmol L−1; estradiol,1.0 μmol L−1

0 2 4 6 8 10 12

30

60

90

120

150

180

c

b

a

Rel

ativ

e C

L In

tens

ity

CH

2O

2

( 10 -2 M)

Fig. 8 Effect of H2O2 concentration on CL intensity. (a) luminol +H2O2 + gold nanoparticles + estrone; (b) luminol + H2O2 + goldnanoparticles + estradiol; (c) luminol + H2O2 + gold nanoparticles +estriol. Conditions: luminol, 0.1 mmol L−1; NaOH, 50 mmol L−1; goldnanoparticles, 50 μmol L−1; estriol, 1.0 μmol L−1; flow rate, 2.8 mLmin−1; estrone, 1.0 μmol L−1; estriol, 1.0 μmol L−1; estradiol,1.0 μmol L−1

Fig. 9 Effect of NaOH concentration on CL intensity. (a) luminol +H2O2 + gold nanoparticles + estrone; (b) luminol + H2O2 + goldnanoparticles + estradiol; (c) luminol + H2O2 + gold nanoparticles +estriol. Conditions: luminol, 0.1 mmol L−1; H2O2, 30 mmol L−1; goldnanoparticles, 50 μmol L−1; estriol, 1.0 μmol L−1; flow rate, 2.8 mLmin−1; estrone, 1.0 μmol L−1; estriol, 1.0 μmol L−1; estradiol,1.0 μmol L−1

590 Anal Bioanal Chem (2007) 387:585–592

reaction. At high flow rates, however, CL emission becameirreproducible, because of turbulence. As a compromisebetween CL emission intensity and reproducibility a flowrate of 2.8 mL min−1 was selected for all experiments.

Effect of the concentration of gold nanoparticles

The effect of the concentration of gold nanoparticles wasalso investigated. The CL intensity increased steadily withincreasing concentration of gold nanoparticles. As acompromise between CL intensity and reagent consump-tion, 50 μmol L−1 gold nanoparticles were used.

Analytical data

Under the optimized experimental conditions the calibra-tion graph for the estrogens showed that the relative CLintensity (ΔI) was linearly proportional to the logarithm ofthe concentration of the estrogen standard solution (C). Thelinear ranges, regression equations, correlation coefficients,and detection limits obtained are summarized in Table 1.The reproducibility of the proposed CL method for assay ofestrogens was evaluated by performing eleven replicateinjections of 0.1 μmol L−1 estrone, 1 μmol L−1 estradiol,and 0.5 μmol L−1 estriol. The relative standard deviationwas 2.9% for 1.0 μmol L−1 estrone, 1.8% for 1.0 μmol L−1

estriol, and 2.6% for 0.7 μmol L−1 estradiol.

Interferences

The effect of some common substances was investigated byanalysis of a standard solution of 1.0 μmol L−1 estriol underthe optimum conditions. The tolerance of each foreignspecies was taken as the largest concentration yielding anerror less than ±5% in the CL signal for estriol. The resultsare listed in Table 2. It can be seen that dopamine,methanol, and some metal ions could interfere with thedetermination. Ghosh et al. [35] have reported that organiccompounds containing hydroxyl (OH), amino (NH2), ormercapto (SH) groups interact readily with gold nano-particles. It is possible that dopamine or methanol interactwith gold nanoparticles in the CL reaction, leading to theobvious change in CL intensity. Many metal ions could, onthe other hand, strongly catalyze the CL reaction of luminoland hydrogen peroxide [30, 31], and could interfere withdetection of estrogen by use of this CL system. Commonlyuse pharmaceutical tablets and healthy human urinecontained very little of these interferents, however, so this

Table 2 Tolerable concentration ratios for some interfering species(<±5% error) in the analysis of 1.0 μmol L−1 estriol

Substance Tolerance concentrationratio

Na+, K+, Cl−, SO2�4 , NO�

3 1000Ca2+, Mg2+ 100Glucose 50Ethanol, uric acid 20Sucrose, dextrin, fructose, lactose 5Ascorbic acid, Cu2+, Fe3+, Cd2+ 0.1Dopamine 0.01

Table 3 Results from determination of estriol in pharmaceuticaltablets (n=5)

Estrioladded(mmolL−1)

Labelclaim(mmolL−1)a

Amountfound(mmolL−1)

Recovery(%)

RSD(%)

ELISAmethod(mmolL−1)b

0.00 0.173 0.159 – 2.3 0.1680.173 0.321 93.1 3.5 –0.346 0.475 91.5 2.7 –

a Per tablet in 100 mL solutionb This ELISA determination was performed, and the data weresupplied, by Yijishan Hospital, Wuhu

Table 1 Linear relationships between chemiluminescence intensityand the concentration of estrogens

Estrogen Linearranges(μmol L−1)

Linearregressionequation(C×10 6

mol L−1)

Correlationcoefficient

Detectionlimit(S/N=3;nmol L−1)

Estrone 0.7–7.0 ΔICL=181.6+61.3lC 0.9908 3.2Estradiol 0.1–10.0 ΔICL=242.6+23.3lC 0.9934 49.0Estriol 0.04–10.0 ΔICL=260.8+25.1lC 0.9952 7.7

Table 4 Results from determination of total estrogens in urinesamples from pregnant women (37–40 weeks gestation)

Sampleno.

Amountdetected(μmolL−1)

Estrioladded(μmolL−1)

Recovery(%)

RSD(%,n=5)

ELISAmethod(μmolL−1)a

1 0.591 0.0 – 1.6 0.5871.0 93.2 2.72.0 101.8 3.2

2 0.783 0.0 – 2.1 0.7911.0 103.2 3.32.0 102.8 3.5

3 1.120 0.0 – 1.9 1.1161.0 95.1 3.42.0 93.7 2.8

a Data supplied by Yijishan Hospital, Wuhu

Anal Bioanal Chem (2007) 387:585–592 591

method can be used for direct determination of theestrogens in tablets or urine sample from pregnant women.

Analytical applications

Determination of estriol in pharmaceutical tablets

A commercially available pharmaceutical tablet was chosento test the proposed method; the results obtained are givenin Table 3. There were no significant differences betweenthe label contents and the results obtained by analysis of thesamples. From the results it is apparent the method issuitable for analysis of real samples.

Determination of estrogen in urine samples

The total concentration of estriol in the urine of normalpregnant women (37–40 weeks gestation) is 0.28–1.23 μmol L−1 and the concentrations of estrone andestradiol are much less [37]. To further test the possibilityof applying this method to real samples, urine samplesobtained from healthy pregnant women were analyzed. Theresults are listed in Table 4. It is apparent the resultsobtained by use of this method correlate well with thosefrom ELISA (Table 4). Relative standard deviations (RSD)ranged from 1.6 to 3.5% for all the urine samples, showingthe method is suitable for clinical analysis.

Conclusions

It has been found that gold nanoparticles catalyze theluminol–H2O2 CL reaction. When estrogens were added tothe gold nanoparticle-catalyzed CL reaction they interactedwith the particles, leading to a substantial reduction in theintensity of the CL. The reaction was combined with flowinjection to establish a novel method for determination ofestrogens on the basis of this inhibition. The method wasused for determination of estriol in pharmaceutical prepara-tions, with satisfactory results. Besides its advantages ofsimplicity, sensitivity, and relatively inexpensive reagents,this work is important for exploring the applications ofnanomaterial-catalyzed liquid-phase CL reaction, and relatedwork is in progress.

Acknowledgements This work was supported financially by theNatural Science Foundation of the Educational Department of AnhuiProvince (Nos. 2006KJ039A and 2006KJ006TD), the NationalNatural Science Foundation of China (No. 20575001), and a Projectfor the Author of Excellent Teacher of Anhui Province.

References

1. Kricka LJ, Thorpe GHG (1983) Analyst 108:1274–12962. Robards K, Worsfold PJ (1992) Anal Chim Acta 266:147–1733. Ding Z, Quinn BM, Haram SK, Pell LE, Korgel BA, Bard AJ

(2002) Science 296:1293–12974. Myung N, Ding Z, Bard AJ (2002) Nano Lett 2:1315–13195. Bae Y, Myung N, Bard AJ (2004) Nano Lett 4:1153–11616. Niazov T, Pavlov V, Xiao Y, Gill R, Willner I (2004) Nano Lett

4:1683–16877. Cui H, Zhang ZF, Shi MJ, Yang X, Wu YL (2005) Anal Chem

77:6402–64068. Zhang ZF, Cui H, Lai CZ, Liu LJ (2005) Anal Chem 77:3324–

33299. Haruta M (1997) Catal Today 36:153–166

10. Wallace WT, Whetten RL (2002) J Am Chem Soc 124:7499–750511. O’Hair RAJ, Khairallah GN (2004) J Cluster Sci 15:331–36312. Prati L, Rossi M (1998) J Catal 176:552–56013. Lopez N, Janssens TVW, Clausen BS, Xu Y, Mavrikakis M,

Bligaard T, Norskov JK (2004) J Catal 223:232–23514. Yang CM, Kalwei M, Schuth F, Chao KJ (2003) Appl Catal A

254:289–29615. Sau TS, Pal A, Pal T (2001) J Phys Chem B 105:9266–927216. Li ZP, Wang YC, Liu CH, Li YK (2005) Anal Chim Acta

551:85–9117. Spierto FW, Gardner F, Smith SJ (2001) Steroids 66:59–6218. Penalver A, Pocurull E, Borrull F, Marce RM (2002) J

Chromatogr A 964:153–16019. Impens S, Courtheyn D, Wasch KD, Brabander HFD (2003) Anal

Chim Acta 483:269–28020. Jin GP, Lin XQ (2005) Electrochim Acta 50:3556–356221. Hu SS, Wu KB, Yi HC, Cui DF (2002) Anal Chim Acta 464:209–

21622. Matsumoto K, Tsukahara Y, Uemura T, Tsunoda K, Kume H,

Kawasaki S, Tadano J, Matsuya T (2002) J Chromatogr B 773:135–142

23. Van den Berg RH, Jansen EHJM, Zomer G, Enkelaar-WillemsenC, Bothmiedema R, Stephany RW (1986) J Chromatogr 360:449–451

24. Zhao LX, Lin JM (2005) J Biotechnol 118:177–18625. Liao SL, Wu XP, Xie ZH (2005) Anal Chim Acta 537:189–19526. Wang J, Ye HZ, Jiang Z, Chen NS, Huang JL (2004) Anal Chim

Acta 508:171–17627. Frens G (1973) Nat Phys Sci 241:20–2228. Wang L, Yang P, Li YX, Zhu CQ (2006) Talanta 70:219–22429. Zhang C, Zhang ZY, Yu BB, Shi JJ, Zhang XR (2002) Anal Chem

74:96–9930. Merényi G, Lind JS, Eriksen TE (1990) J Biolumin Chemilumin

5:53–5631. Burdo TG, Seitz WR (1975) Anal Chem 47:1639–164332. Li Y, Zhu C, Wang L (2005) Microchim Acta 150:95–9933. Merényi G, Lind JS (1980) J Am Chem Soc 102:5830–583534. Kcicka LJ, Voyta JC, Bronstein I (2000) Methods Enzymol

305:370–39035. Ghosh SK, Nath S, Kundu S, Esumi K, Pal T (2004) J Phys Chem

B 108:13963–1397136. Vijay Sarathy K, Kulkarni GU, Rao CNR (1997) Chem Commun

537–53837. Mei ZW, Pan BS (2001) Hand book of recently clinical detection.

Hunan Science and Technology Press, Changsha p 251

592 Anal Bioanal Chem (2007) 387:585–592