PAPER www.rsc.org/analyst | Analyst

Publ

ishe

d on

29

July

201

0. D

ownl

oade

d by

Uni

vers

ity o

f M

ichi

gan

Lib

rary

on

21/1

0/20

14 2

1:37

:52.

View Article Online / Journal Homepage / Table of Contents for this issue

Sulfide sensor based on the room-temperature phosphorescence of ZnO/SiO2

nanocomposite†

Na Wang,a Ting Zhou,b Jun Wang,a Hongyan Yuan*a and Dan Xiao*ab

Received 11th February 2010, Accepted 22nd June 2010

DOI: 10.1039/c0an00081g

An environmentally friendly sulfide sensor based on the room-temperature phosphorescence (RTP) of

the ZnO/SiO2 nanocomposite has been developed. The ZnO/SiO2 nanocomposite prepared by the

sol–gel route produces highly emissive broadband RTP which can be clearly observed by the naked eye.

The phosphorescence intensity monitored at 460 nm (excitation at 320 nm) decreases with increasing

sulfide ions concentrations. The response behavior of the sensor is dependent on the pH value of the

solution. At pH 10, the sensor shows a good, linear response to sulfide from 4.88 � 10�5 to 1.02 � 10�2

M with a detection limit of 1.64 � 10�6 M (3s). It has been successfully applied to the determination of

sulfide in spiked water and wastewater. Furthermore, this sensor can be regenerated by dipping it into

an H2O2 solution. The mechanisms for the RTP detection of sulfide based on the ZnO/SiO2

nanocomposite and the sensor regeneration by H2O2 are proposed.

Introduction

Sulfide anion exists widely in industrial locations such as

tanneries, food processing plants, petroleum refineries, paper and

pulp manufacturing plants, where it is either used as a reactant or

is produced as a by-product of manufacturing or industrial

processes.1,2 Sulfide is a highly undesirable contaminant in

effluents because of its high toxicity to living organisms, its

capacity to remove dissolved oxygen and its capability to

produce hydrogen sulfide.3 Additionally, numerous problems

including the corrosion of metal surfaces and the degradation of

concrete are related to sulfide build-up.3,4 The toxic and corrosive

nature of sulfide makes it very important to detect sulfide ions in

effluents. Various methods have been employed in the detection

of sulfide such as titration,5 spectroscopy,1,2,4–8 electro-

chemistry,3,6,9 chromatography10 and combinations thereof.11,12

Among these methods, optical sensing approaches have attracted

the attention of many researchers with respect to the cost, ease of

preparation, insensitivity to electrical interferences and the

possibility of remote sensing of hazardous areas using optical

fibre techniques. The reported optical sensing approaches are

sensitive in detecting sulfide; however, the sensors based on the

methylene blue (MB) method cannot be regenerated5,6,11 and the

autofluorescence and scattering light from matrices usually

interfere with determination using fluorescence-based

methods.2,4,7 Moreover, one major disadvantage of these

approaches is that some reagents used to develop sensors are

aCollege of Chemical Engineering, Sichuan University, Chengdu 610065,P. R. China. E-mail: yuan_hy@scu.edu.cn; xiaodan@scu.edu.cn; Fax:+86-028-85416029; Tel: +86-028-85415029bKey laboratory of Green Chemistry and Technology, Ministry ofEducation, College of Chemistry, Sichuan University, Chengdu 610065,P. R. China. E-mail: xiaodan@scu.edu.cn; Fax: +86-028-85416029; Tel:+86-028-85415029

† Electronic supplementary information (ESI) available: The details ofselection of preparation conditions and additional information ofcharacterization and performance of regenerated sensor. See DOI:10.1039/c0an00081g

2386 | Analyst, 2010, 135, 2386–2393

toxic. The MB test is the most common approach to the analysis

of sulfide.5,6,11 But the N-dimethyl-p-phenylenediamine (DMPD)

is a poisoning and cancer-causing reagent.13 And the commonly

used fluorescent sensing systems contain Hg or Cd complexes,2,4,7

which are one of the most hazardous materials.13 Recently, we

first reported the sulfide sensor based on the room-temperature

phosphorescence of the PbO/SiO2 nanocomposite.8 But lead and

its compounds are highly toxic when eaten or inhaled.8,13 These

optical sensors pose additional pollution problems for the envi-

ronment in the preparation and detection processes, thus an

environmentally friendly optical sensing system for the deter-

mination of sulfide is strongly desirable.

On the other hand, as a very useful mode of detection for

optical sensing applications,14 the advantages of room-tempera-

ture phosphorescence (RTP) detection over the fluorescence

method are longer lifetimes of emission, the larger Stokes’ shifts

and the superior selectivity.15 The long lifetime of phosphore-

scence can avoid interferences from any short-lived fluorescence

emission and scattering light by simply setting an appropriate

delay time.15,16 The selectivity is superior because the phospho-

rescence is a less usual phenomenon.15 Recently, the RTP of

doped semiconductors has attracted considerable interest.8,16,17

The sensors based on the RTP of PbO/SiO2,8 Mn-ZnS/SiO216 and

TiO2/SiO217 have been successfully employed in detecting the

chemical species. To the best of our knowledge, however, the

optical sensor based on the RTP of ZnO/SiO2 has never been

reported.

In this paper, we report an environmentally friendly optical

sensor for the determination of sulfide based on the RTP of the

ZnO/SiO2 nanocomposite. We prepared the ZnO/SiO2 nano-

composite via the one-pot sol–gel route. The reagents used in the

experiment have low toxicity13 and the preparation process does

not produce wastewater. In addition, the ZnO/SiO2 nano-

composite is innocuous and easily separated away from water.18

This material shows the afterglow phenomenon with a long

phosphorescence lifetime. Its phosphorescence intensity is

remarkably quenched by sulfide without interferences from

This journal is ª The Royal Society of Chemistry 2010

Publ

ishe

d on

29

July

201

0. D

ownl

oade

d by

Uni

vers

ity o

f M

ichi

gan

Lib

rary

on

21/1

0/20

14 2

1:37

:52.

View Article Online

common ions and the scattering light of matrices. Moreover the

sensor can be regenerated by dipping it into an H2O2 solution.

The origins of RTP for the ZnO/SiO2 nanocomposite, the

mechanisms for the RTP detection of sulfide based on the ZnO/

SiO2 nanocomposite and the sensor regeneration by H2O2 are

discussed in detail. This sensor offers the advantages of non-

toxicity, simplicity, selectively, reproducibility and good

stability, which make it a valuable alternative to other optical

sensing systems employed to date for the detection of sulfide.

Experimental

Reagents

Zinc acetate dihydrate was purchased from Shanghai Chemicals

(Shanghai, China). Tetraethoxysilane, acetic acid and ethanol

were purchased from Tianjin Chemicals (Tianjin, China).

Sodium sulfide nonahydrate, nitric acid, ammonium hydroxide

and 30% hydrogen peroxide aqueous solution were obtained

from Chengdu Chemicals (Sichuan, China). Note: sodium sulfide

is a corrosive reagent and must be handled with extreme caution.

Do not inhale vapors and prevent being injured!3 All reagents

were of analytical reagent grade and were used without further

purification. The standard Na2S solutions were freshly prepared

in 50 mM phosphate buffers at pH 6.0 to 11.0 and initially

standardized by iodometric titration.5 Redistilled water was used

throughout all experiments.

Instrumentation

All the phosphorescence spectra were obtained from a fluore-

scence spectrophotometer F-4500 (Hitachi, Japan) at room

temperature. Scanning electron microscope (SEM) analysis was

performed on a VEGA-II (Tescan, Czech Republic) system

equipped with energy-dispersive X-ray spectroscopy (EDS).

Transmission electron microscopy (TEM) images were taken

using an H-800 electron microscope and H-8010 scanning system

(Hitachi, Japan). The X-ray diffraction (XRD) patterns were

obtained using a DX-1000 powder XRD (Dandong Fangyuan,

China) with Cu Ka radiation. Fourier transform infrared (FT-

IR) spectra were performed on a 670 FT-IR spectrophotometer

(Thermo Nicolet, USA) with KBr discs. X-Ray photoelectron

spectroscopy (XPS) analysis was performed on an XSAM800

(Kratos, UK) XPS instrument. The X-ray fluorescence spec-

troscopy (XRF) analysis was carried out using an XRD-1800

sequential XRF spectrometer (Shimadzu, Japan). The pH

measurements were taken on an Orion 720+ combined pH glass

electrode (Thermo Electron, USA). The flow rate of N2 was

controlled using mass flow controllers (Brooks, The Nether-

lands). A ZF5 UV lamp (Shanghai Jiapeng, China) was used for

UV irradiation when taking the RTP photographs. And a tablet

press machine DF-4 (Tianjin Gangdong, China) was used to

press the ZnO/SiO2 powder into a disk; compressive stress was

set at 8 MPa.

Preparation of the ZnO/SiO2 nanoparticles

The nanometer particles of ZnO/SiO2 were prepared by the

sol–gel route. In a typical procedure, 6.00 mL tetraethoxysilane

and 3.00 mL acetic acid were dissolved in 10.0 mL ethanol and

This journal is ª The Royal Society of Chemistry 2010

the mixed solution was stirred for 15 min. Then 3.00 mL redis-

tilled water, 1.00 mL nitric acid and 0.6553 g zinc acetate

dehydrate were sequentially added to the solution under

magnetic stirring at intervals of 15 min (Zn/Si molar ratio ¼10/90). After 30 min the zinc acetate dihydrate was dissolved

completely, and 2.00 mL of 9% ammonium hydroxide was added

to the solution. The mixed solution was stopped stirring until it

became a transparent gel. Then the gel was allowed to age for

24 h. All of the experiments were performed at room tempera-

ture. Finally, the gel was calcined at 500 �C for 2 h under a static

air atmosphere. The resultant ZnO/SiO2 nanoparticles were

ground manually into a powder and pressed into disks of 12 mm

diameter and 1 mm thickness. These disks were kept in

a desiccator before use. This preparation method has good

reproducibility. When nine batches of materials were prepared

according to above process, the RSD of RTP intensity was

2.77%. For comparison, an SiO2 sample was prepared by the

same route as for the ZnO/SiO2 nanocomposite, except that no

zinc acetate dihydrate was added.

Measurement procedure

Phosphorescence measurements of the sensor were carried out on

a fluorescence-phosphorescence spectrophotometer in the

absence and presence of a series of Na2S solutions when the

spectrophotometer was set in the phosphorescence mode with

chopper arrangement. The slit widths of excitation and emission

were 5 and 10 nm, respectively. The scan speed was 240 nm/min.

The gate time and delay time were the default values of the

Hitachi F-4500 instrument. The photomultiplier tube (PMT)

voltage was set at 750 V. The response of the sensor to Na2S

solutions of different concentrations was detected using a regular

quartz cuvette as shown in Fig. S1 in the ESI† according to our

previous work.8 After ZnO/SiO2 interaction with sulfide for

600 s, the wavelength scan phosphorescence spectra were

obtained with the maximum excitation and emission wavelength

at 320 nm and 460 nm, respectively.

Results and discussion

Selection of preparation conditions and characterization of the

ZnO/SiO2 nanocomposite before and after addition of Na2S

The phosphorescence intensity and experimental conditions are

summarized in Table S1 and Fig. S2 (ESI†). The Zn/Si molar

ratio and also the calcination temperature and time have an

effect on the phosphorescence intensity and spectral character-

istics. From the results listed in Table S1, it is found that the long

time calcinations or high-temperature calcinations result in

a large blue shift of the excitation and emission spectra.19 The

XRD spectra of these materials do not provide any apparent

information about the changes of phosphorescence intensity and

spectra; thus, we ascribe these changes to the competitive results

of two radiative transitions (Fig. S3 (ESI†)). In view of the

phosphorescence intensity and the response behavior for sulfide

sensing (Table S1 and Fig. S4 (ESI†)), the Zn/Si molar ratio of

10/90 and calcination performed at 500 �C for 2 h was selected as

the experimental conditions.

The morphological studies and elementary distributions of the

material studies were carried out by SEM, TEM and EDS as

Analyst, 2010, 135, 2386–2393 | 2387

Fig. 1 The (A) XRD and (B) XRF spectra of the as-prepared ZnO/SiO2

nanocomposite sample (a) before and (b) after interaction with Na2S.

Fig. 2 XPS spectra of the (A) Zn 2p3/2 and (B) Si 2p of ZnO/SiO2

nanocomposite (a) before, (b) after interaction with Na2S and (c) after

being regenerated with H2O2.

Publ

ishe

d on

29

July

201

0. D

ownl

oade

d by

Uni

vers

ity o

f M

ichi

gan

Lib

rary

on

21/1

0/20

14 2

1:37

:52.

View Article Online

shown in Fig. S5 and Fig. S6 (ESI†). The SEM and TEM images

(Fig. S5 and Fig. S6) show that the ZnO/SiO2 nanoparticles have

a wide size distribution due to aggregation. The diameter of

particles was estimated to be between 20 and 50 nm. The EDS

images in Fig. S6 show the Si, O and Zn elements to be equally

distributed within the material. However, the sulfur signal was

only found in the ZnO/SiO2 nanocomposite after interaction

with Na2S (Fig. S6(b)). Fig. 1 illustrates the XRD patterns and

XRF spectra of the ZnO/SiO2 nanocomposite before and after

the addition of Na2S. In Fig. 1A(a), the wide angle XRD patterns

show only the diffraction peak of non-crystalline silica frame-

works, and no characteristic peaks belonging to ZnO have been

observed in the ZnO/SiO2 nanocomposite. The very weak peaks

of ZnO attributed to its nature and/or to its small amount20 may

be concealed by the strong and broad peak of amorphous SiO2.

However, after ZnO/SiO2 interaction with sulfide, the XRD

profile shows the presence of the weak diffraction peaks, whose

positions fully match those of the ZnS blende crystal structure as

shown in Fig. 1A(b). It demonstrates that ZnS forms after the

ZnO/SiO2 interaction with Na2S. As shown in Fig. 1B, the S Ka

line can only be observed in the XRF spectrum for the ZnO/SiO2

nanocomposite after interaction with Na2S, which indicates that

sulfide is present in the quenched ZnO/SiO2 composite oxides.

An XPS study has been carried out on the ZnO/SiO2 nano-

composite before and after being quenched so as to obtain

additional information about the phosphorescent material. The

result of XRD experiments did not apparently show the existence

of ZnO in the as-prepared nanocomposite oxide; however, XPS

results did prove the existence of ZnO in the nanocomposite. As

shown in Fig. 2A(a), the binding energy of Zn 2p3/2 in the ZnO/

SiO2 is 1022.4 eV, which is 1.3 eV higher than that in pure

2388 | Analyst, 2010, 135, 2386–2393

ZnO.21,22 Due to the higher electronegativity of Si than that of

Zn, the valence electron density of Zn in the Zn–O–Si bond is

lower than that in the Zn–O–Zn bond, which makes it reasonable

to ascribe the shift of binding energy of Zn 2p3/2 to the formation

of Zn–O–Si.21,22 Because the replacement of silicon by divalent

metal cations in a silica network is thermodynamically

unfavorable,23 the Zn–O–Si bonds should exist on the interface

between the ZnO particles and the pore walls of SiO2 resulting in

an increase in covalence with respect to bulk ZnO.21,22 After

interaction with Na2S, the Zn 2p3/2 and Si 2p peaks of the ZnO/

SiO2 nanocomposite shift to a lower binding energy. It indicates

the changes in the chemical environment of the ZnO/SiO2

nanocomposite and that ZnS/SiO2 forms after interaction with

Na2S.8 And after being regenerated with H2O2, the binding

energy values of Zn 2p3/2 and Si 2p are the same as those of ZnO/

SiO2, according to which we suppose that after treatment with

H2O2 the ZnS in the nanocomposite of ZnS/SiO2 is oxidized to

ZnO. Additionally, the FT-IR spectra of the SiO2, ZnO/SiO2 and

ZnS/SiO2 are shown in Fig. S7 (ESI†). The shifts of these peaks

in the FT-IR spectra illustrate the changes in the chemical

environment of the material.8

Effect of pH on the performance of sulfide sensor based on ZnO/

SiO2 nanocomposite

The pH of the buffer solutions is selected considering the

following two aspects: (1) the stability of ZnO/SiO2; and (2) the

suitable conditions for phosphorescence measurement. From

This journal is ª The Royal Society of Chemistry 2010

Publ

ishe

d on

29

July

201

0. D

ownl

oade

d by

Uni

vers

ity o

f M

ichi

gan

Lib

rary

on

21/1

0/20

14 2

1:37

:52.

View Article Online

the results of the experiments (data not given), it was observed

that the ZnO/SiO2 nanocomposite was unstable at either pH

below 6.0 or pH above 10.0. This is because ZnO is an ampho-

teric oxide and the structure will collapse in both strong acidic

and alkali solution. In order to maintain the stabilization of the

ZnO/SiO2 nanocomposite, the pH dependence of the phospho-

rescence intensity of the phosphor was tested over the range of

pH from 6.0 to 10.0. If the phosphorescence intensity is

proportional to the concentration of the phosphor, then the

value of phosphor response, a, can be represented as: a ¼ (P0 �P)/P0, where P0 and P are phosphorescence intensities in the

absence and in the presence of the sulfide, respectively. In Fig. S8

(ESI†), the phosphor response, a, is plotted as a function of�lgC

at pH 6.0–10.0. The parameter C is the concentration of Na2S

solution (M). The response behavior of this sensor is pH

dependent. The phosphor has lower a values at pH 6.0 and 7.0

than that at pH values from 8.0 to 10.0. When the pH value

decreases, the concentration of hydrosulfide ions decreases with

the partial conversion of hydrogen sulfide molecules. Actually,

sulfide ion is seldom present in aqueous solutions and the only

form present at pH values higher than 9 is HS�.5,12 As a result, the

greatest response of the sensor to Na2S occurs at pH 10.0. Thus,

pH 10.0 is chosen as the working condition.

Performance of ZnO/SiO2 nanocomposite for Na2S

determination

To evaluate the sensitivity of the ZnO/SiO2 nanocomposite for

the determination of sulfide, the intensity change of the ZnO/

SiO2 phosphorescence was monitored at 460 nm. The response of

the sensor at increasing Na2S concentration is shown in Fig. 3.

Upon the addition of different concentrations of Na2S, the

phosphorescence intensity of the ZnO/SiO2 nanocomposite

decreased. The phosphorescence intensity of the ZnO/SiO2

nanocomposite and the concentrations of Na2S accord with the

logarithmic quantitative equation, lg[(P0 � P)/P] ¼ a + blgC, in

which P0 is the original phosphorescence intensity of the ZnO/

SiO2 nanocomposite, P is the phosphorescence intensity of the

Fig. 3 The effect of Na2S concentration on the emission spectra of ZnO/

SiO2 nanocomposite and its calibration curves. The sulfide concentration:

(a) 0, (b) 4.88 � 10�5, (c) 1.14 � 10�4, (d) 2.44 � 10�4, (e) 4.39 � 10�4, (f)

1.09 � 10�3, (g) 1.42 � 10�3, (h) 2.07 � 10�3, (i) 3.69 � 10�3, (j) 5.32 �10�3, (k) 6.95 � 10�3, (l) 8.57 � 10�3 and (m) 1.02 � 10�2 M.

This journal is ª The Royal Society of Chemistry 2010

ZnO/SiO2 nanocomposite in different concentrations of Na2S

solution, and C is the concentration of Na2S solution (M). The

calibration curve of the sensor is shown as an inset in Fig. 3. The

linear equation is y ¼ 1.1535x � 3.1412, in which y is for �lg[(P0

� P)/P] and x is for�lgC. The linear response range of the sensor

to Na2S concentrations is from 4.88 � 10�5 to 1.02 � 10�2 M

(linear related coefficient R2 ¼ 0.9914). The detection limit,

calculated as three times the standard deviation of the blank

signal, is 1.64 � 10�6 M. The t95 values (time to reach 95% of

maximum response) were <600 s. The sensing time was inde-

pendent of sulfide concentration because measurements were

taken upon the steady state being reached.

In view of the reduction of sulfide, the phosphorescence

intensity may be recovered by some oxidizing agents. In this

work, we found that hydrogen peroxide can recover the phos-

phorescence intensity of the ZnO/SiO2 nanocomposite. After

immersion of a quenched sensor into a 0.01 M H2O2 solution for

4 h, 95.2% phosphorescence intensity was found to be recovered

(Fig. S9 (ESI†)). Upon the addition of different concentrations

of Na2S, the phosphorescence intensity of the sensor after being

regenerated once became lower and lower as shown in Fig. S10

(ESI†). The regenerated sensor shows a good, linear response to

sulfide from 4.88 � 10�5 to 1.02 � 10�2 M. The linear equation is

y ¼ 1.135x � 2.7215 (R2 ¼ 0.9933). After the sensor was

regenerated twice according to the process above, just 81.8%

phosphorescence intensity was recovered (Fig. S9 (ESI†)) and it

showed a narrower linear response to sulfide from 2.44 � 10�4 to

1.02 � 10�2 M with the detection limit of 1.13 � 10�5 M. After

repeated experiments, some of the sodium ions and phosphate

ions were absorbed on the surface of ZnO/SiO2 and could not be

completely eluted which thus gradually passivated the phos-

phorescence centers. It may be partly responsible for both the

decrease of recovered phosphorescence intensity and the

performance of the regenerated sensor.

Although the color of the ZnO/SiO2 nanocomposite did not

change when different concentrations of Na2S solution were

added, the change of phosphorescence intensity could be

observed by the naked eye (Fig. 4). Fig. 4B shows the RTP

photograph of the sensor in sulfide concentrations ranging from

0 to 1.00 � 10�2 M, captured with a commercial digital camera,

right after turning off the 254 nm UV lamp (ESI†). It can be seen

that the phosphorescence intensity of the sensor decreases with

increasing Na2S concentrations. The result indicates that the

sensor provides a simple optical method for sulfide analysis.

Fig. 4 (A) Photograph and (B) RTP photograph of the ZnO/SiO2

nanocomposite upon excitation at 254 nm at Na2S concentrations of (a)

0, (b) 2.00 � 10�3, (c) 6.00 � 10�3 and (d) 1.00 � 10�2 M.

Analyst, 2010, 135, 2386–2393 | 2389

Publ

ishe

d on

29

July

201

0. D

ownl

oade

d by

Uni

vers

ity o

f M

ichi

gan

Lib

rary

on

21/1

0/20

14 2

1:37

:52.

View Article Online

The proposed mechanisms for the RTP detection of sulfide based

on the ZnO/SiO2 nanocomposite and the sensor regeneration by

H2O2

In order to understand the mechanisms for the RTP detection of

sulfide based on the ZnO/SiO2 nanocomposite and the sensor

regeneration by H2O2 it is important to find out the origin of the

RTP for the ZnO/SiO2 nanocomposite. As shown in Fig. 5A(a1)

and (a2), the maximum excitation and emission wavelengths of

the ZnO/SiO2 nanocomposite are 320 nm and 460 nm, respec-

tively. The phosphorescence intensity monitored at 460 nm

(excitation at 320 nm) decreases obviously with the addition of

1.00 � 10�2 M sulfide (Fig. 5A(a3)). To our surprise, however,

when excited at 350 nm, the phosphorescence intensities of the

ZnO/SiO2 nanocomposite monitored at 490 nm in the absence

and in the presence of 1.00 � 10�2 M sulfide are 299 and 285,

respectively (Fig. 5A(b1) and (b3)). The nearly unchanged

phosphorescence intensity (decreased ca. 4.7%) illustrates the

emission spectrum peaking at 490 nm upon excitation at 350 nm

is almost unaffected by the sulfide. Its emission intensity is

apparently weaker than that by excitation at a wavelength of

320 nm (Fig. 5A(b1) and (a2)). After interaction with sulfide, the

phosphorescence intensity at the emission wavelength of 460 nm

is almost completely quenched (Fig. 5A(a3)) and the excitation

and emission wavelengths of the phosphorescent material change

Fig. 5 (A) The room-temperature phosphorescence (a1) excitation and (a2) em

(a3) emission spectrum monitored at 460 nm (excitation at 320 nm) of ZnO

emission spectrum monitored at 490 nm (excitation at 350 nm) of the ZnO

phosphorescence (b2) excitation and (b3) emission spectra of the ZnO/SiO2 na

temperature phosphorescence emission spectra of (a) ZnO/SiO2 nanocompo

350 nm; inset is a close-up view of the spectrum for (b). (C) Gaussian fitted

nanocomposite at 460 nm upon excitation at 320 nm and (D) the energy leve

2390 | Analyst, 2010, 135, 2386–2393

to 350 nm and 490 nm, respectively (Fig. 5A(b2) and (b3)).

Considering that the SiO2 does not react with sulfide, we suppose

that the emission band peaking at 490 nm originates from the

SiO2. To confirm such a supposition we studied the emission

spectrum of as-prepared SiO2 by excitation at a wavelength of

350 nm (Fig. 5B). It is clear from Fig. 5B that the emission

wavelength of SiO2 upon excitation at 350 nm is at 490 nm, which

is identical to that of ZnO/SiO2 upon excitation at 350 nm with

just a decrease in intensity. The enhancement in the RTP inten-

sity of SiO2 in ZnO/SiO2 nanocomposite may be caused by the

effect of the interface and the nanocomposite structure.21 Thus,

we ascribe the emission peak for the ZnO/SiO2 nanocomposite at

490 nm upon excitation at 350 nm to the defects of SiO224 while

we ascribe the emission peak at 460 nm upon excitation at

320 nm to the defects of ZnO. However, the spectrum of the

maximum emission wavelength at 490 nm does not affect the

sulfide determination performance of the sensor since the 30 nm

interval of emission wavelength is wide enough to eliminate the

interference.

The emission peak at 460 nm of ZnO is asymmetric, indicating

the presence of more than one peak. To analyze the exact origin

leading to the RTP emission of ZnO at 460 nm, the RTP spectra

were further studied using Gaussian fitting. As shown in Fig. 5C,

the emission spectrum at 460 nm is nicely fitted by three peaks

ission spectra of the ZnO/SiO2 nanocomposite before addition of sulfide;

/SiO2 nanocomposite after interaction with 1.00 � 10�2 M sulfide; (b1)

/SiO2 nanocomposite before addition of sulfide; the room-temperature

nocomposite after interaction with 1.00 � 10�2 M sulfide. (B) The room-

site and (b) as-prepared SiO2 monitored at 490 nm upon excitation at

peaks of room-temperature phosphorescence emission of the ZnO/SiO2

l diagram of ZnO; red short-dash lines are the three origins of RTP.

This journal is ª The Royal Society of Chemistry 2010

Publ

ishe

d on

29

July

201

0. D

ownl

oade

d by

Uni

vers

ity o

f M

ichi

gan

Lib

rary

on

21/1

0/20

14 2

1:37

:52.

View Article Online

centered at 460 nm (�2.7 eV), 452 nm (�2.74 eV) and 515 nm

(�2.41 eV). They are characteristic emission bands of ZnO25–28

which further confirm the emission band centered at 460 nm

originate from ZnO. As in the energy level diagram of ZnO

shown in Fig. 5D, the peak at 460 nm is related to the anion

vacancies, which lies 2.7 eV below the bottom of the conduction

band in ZnO;25 the peak at 452 nm is related to the interstitial

zinc;26 and the peak at 515 nm is associated with the single

ionized charge state of the single ionized oxygen vacancy27 or the

antisite defects (OZn).28 According to the report that surface

defects can act as traps for phosphorescence,29 it can thus be

deduced that there should exist plenty of oxygen vacancies,

interstitial zinc and antisite defects on the surface of the ZnO/

SiO2 nanocomposite, which might be the reason for the obser-

vation of the afterglow phenomenon.30

The phosphorescence quenching is usually divided into static

quenching and dynamic quenching. The principle of the

quenching was investigated by measuring the phosphorescence

lifetimes of the ZnO/SiO2 nanocomposite before and after

interaction with Na2S. The decay curves are shown in Fig. S11

(ESI†). According to the calculation method of previous

reports,17 the phosphorescence lifetimes of the sensor in the

absence and in the presence of Na2S are both 0.55 s, suggesting

that the quenching process is a static quenching. The result

implies that after the addition of Na2S, a non-phosphorescence

ground state complex ZnS formed between the surface of ZnO

and Na2S, as supported by the XRD (Fig. 1A) and XPS (Fig. 2).

Although ZnS is a well-known phosphor, only wurtzite ZnS

rather than cubic zinc blende doped with Cu2+ or Eu2+ can show

a relatively good afterglow performance.31

In our previous work, the chemical reaction of sensor

regeneration by H2O2 has been studied.8 And in this study, the

sensor can also be regenerated with the treatment of simply

immersing the sensor into H2O2 solution. We suppose that ZnS

Fig. 6 Schematic illustrations for mechanisms of detection of sulfide

based on the RTP of ZnO/SiO2 nanocomposite and the recovery of

phosphorescence intensity by the H2O2.

This journal is ª The Royal Society of Chemistry 2010

reacts with H2O2 to form sulfate and ZnO, which results in the

recovery of the phosphorescence intensity. To confirm the

formation of sulfate, the quenched sensor after being washed

with redistilled water repeatedly was put into an H2O2 solution,

and then BaCl2 was added to the H2O2 solution. The white

muddle appeared immediately, which illustrates that the

precipitation of BaSO4 was produced. It indicated the formation

of sulfate after the ZnS/SiO2 being oxidized by H2O2. The

proposed chemical reactions of the sensor with sulfide ions and

the regeneration of the sensor by H2O2 are shown in Fig. 6.

Moreover, in view of the phosphorescence originating mainly

from the oxygen vacancies (Fig. 5C and 5D) of ZnO, although

the quenched sensor (i.e. ZnS/SiO2) can be oxidized to ZnO/SiO2

by H2O2, the surface defects of ZnO especially the oxygen

vacancies are also affected by the oxidizing agents during the

regeneration process and cannot be totally regenerated. As

a result, the number of surface defects decreases obviously after

two regenerations, which provides another explanation for both

the decrease of the recovered phosphorescence intensity and the

performance of the regenerated sensor.

Application

In order to assess the reliability and accuracy of the proposed

method, the sensor was applied to the determination of sulfide in

water samples: spiked water and wastewater collected in a local

leather plant. As the interference from the wastewater turbidity

hinders spectroscopic evaluation, the wastewater samples were

pretreated by the method of gas-phase separation and sorption32

and diluted 100 times before direct analysis. They showed that all

sulfide in the wastewater samples could be extracted by this

method. The results of the proposed method were found to agree

reasonably with the results of the MB test5 (Table 1). The

quantitative recovery, from 92.8 to 103%, demonstrates the

reliability of the proposed method. In order to assess the reli-

ability of quantitative measurement for the regenerated sensor,

the sensor after being regenerated once was used to determine

sulfide in the leather wastewater (Table 1, Wastewater 2b). The

results illustrate that the sensor shows good and precise results

for the determination of sulfide in real samples.

Interference studies

The tolerance ratios (Cchemical species/Csulfide) when interference

concentration is varying the analytical signal by 5% are present

as follows: Na+, K+, Cl�, I�, PO43�, HPO4

2�, H2PO4�, SCN�,

CN�, methanol, ethanol, acetone, dichloromethane, chloroform,

benzene and dimethyl sulfoxide (1000); SO42�, NH4

+ (200);

CH3COOH, Br�, S2O32�, SO3

2�, HSO3�, C2O4

2� (100); CO32�

(50); NO3�, HCO3

�, F�, NO2� (40); phenol, mercaptopyridine

(10). The phenol and mercaptopyridine interfere with the deter-

mination of sulfide seriously. This may be caused by the strong

binding strength between the hydroxyl of the phenol or the

mercapto of mercaptopyridine and ZnO. And other chemical

species do not interfere much.

Lifetime of the sensor

To demonstrate the stability of the sensor, five and a half months

later, the proposed sensor stored in a desiccator was used to

Analyst, 2010, 135, 2386–2393 | 2391

Table 1 Determination of sulfide in spiked water and wastewater samples and recovery test

Sample Added (�10�5 M)Proposed methodfound (�10�5 M)a Recovery (%) RSD (%)

MB methodfound (�10�5 M) Relative error (%)

Spiked water 9.76 10.1 103 0.898 9.80 +2.9748.8 48.4 97.8 0.512 49.5 �2.27

Wastewater 1b 0 11.3 — 0.704 12.0 �6.1919.5 31.5 102 0.720 — —

Wastewater 2b 0 12.5 — 1.937 12.0 +4.0019.5 29.7 92.8 1.384 — —

a Average value of five measurements. b Wastewater 1 and wastewater 2 are the same leather wastewater samples. Wastewater 1 contained sulfide andwas detected by the sensor and wastewater 2 contained sulfide and was detected by the sensor regenerated once.

Publ

ishe

d on

29

July

201

0. D

ownl

oade

d by

Uni

vers

ity o

f M

ichi

gan

Lib

rary

on

21/1

0/20

14 2

1:37

:52.

View Article Online

detect the sulfide concentrations according to the process above.

The wavelength spectra of the sensor are shown in Fig. S12

(ESI†). The original phosphorescence intensity of the sensor

decreased from 766.8 to 762.6 (decreased ca. 0.55%). Upon the

addition of different concentrations of sulfide (from 4.88 � 10�5

to 1.02 � 10�2 M), the phosphorescence intensity gradually

decreased. The linear equation of this is y ¼ 0.9983x � 2.3017.

The linear-related coefficient of the sensor is R2 ¼ 0.9904. The

result demonstrates that the sensor stored in a desiccator shows

satisfactory stability and lifetime.

Conclusions

An environmentally friendly optical sensor for sulfide deter-

mination based on the RTP of the ZnO/SiO2 nanocomposite has

been fabricated and successfully applied to the determination of

sulfide in spiked water and wastewater. The emission band

centered at 460 nm originates from ZnO while that centered at

490 nm originates from SiO2. The phosphorescence of ZnO can

be quenched by the sulfide. At pH 10.0, the sensor exhibits good

linear response to sulfide. Besides interferences from common

chemical species, autofluorescence and the scattering light of the

matrix are avoided in the present sensor based on the RTP

method. And the sensor can be regenerated by H2O2. Further-

more, it has good stability, just needing to be kept in a desiccator.

The results show that this research will provide a new strategy for

the environmentally friendly detection of sulfide in real sample

analysis applications.

Acknowledgements

We thank the financial supports from the National Natural

Science Foundation of China (No. 20775050). We also thank

Prof. Hong Chen and Xi Wu in Analytical & Testing Center of

Sichuan University for the measurement of XPS and XRF,

respectively.

References

1 X. F. Yang, L. P. Wang, H. M. Xu and M. L. Zhao, Anal. Chim. Acta,2009, 631, 91.

2 A. E. Eroglu, M. Volkan and O. Y. Ataman, Talanta, 2000, 53, 89.3 M. I. Prodromidis, P. G. Veitsistas and M. I. Karayannis, Anal.

Chem., 2000, 72, 3995.4 M. M. F. Choi, Analyst, 1998, 123, 1631.5 M. J. Taras, A. E. Greenberg, R. D. Hoak and M. C. Rand, ed., in

Standard Methods for the Examination of Water and Wastewater,

2392 | Analyst, 2010, 135, 2386–2393

American Public Health Association, Washington, D.C., 13th edn,1971, pp. 336–337 and 551–559.

6 N. S. Lawrence, J. Davis and R. G. Compton, Talanta, 2000, 52, 771.7 M. M. F. Choi and P. Hawkins, Sens. Actuators, B, 2003, 90, 211;

M. M. F. Choi and P. Hawkins, Anal. Chim. Acta, 1997, 344, 105;J. Rodr�ıguez-Fern�andez, J. M. Costa and R. Pereiro, Anal. Chim.Acta, 1999, 398, 23.

8 T. Zhou, N. Wang, C. H. Li, H. Y. Yuan and D. Xiao, Anal. Chem.,2010, 82, 1705.

9 N. S. Lawrence, R. P. Deo and J. Wang, Anal. Chim. Acta, 2004, 517,131; D. Giovanelli, N. S. Lawrence, L. Jiang, T. G. J. Jones andR. G. Compton, Analyst, 2003, 128, 173; D. Giovanelli,N. S. Lawrence, L. Jiang, T. G. J. Jones and R. G. Compton, Sens.Actuators, B, 2003, 88, 320.

10 D. Tang and P. H. Santschi, J. Chromatogr., A, 2000, 883, 305.11 S. S. M. Hassan, S. A. M. Marzouk and H. E. M. Sayour, Anal. Chim.

Acta, 2002, 466, 47.12 C. Giuriati, S. Cavalli, A. Gorni, D. Badocco and P. Pastore,

J. Chromatogr., A, 2004, 1023, 105.13 P. Patnaik, in A comprehensive guide to the hazardous properties of

chemical substances, Wiley & Sons, Inc., Hoboken, New Jersey, 3rdedn, 2007, part B, pp. 105–663.

14 C. Chi, C. Im and G. Wegner, J. Chem. Phys., 2006, 124, 024907(1);Y. He, H. F. Wang and X. P. Yan, Chem.–Eur. J., 2009, 15, 5436;A. L. Thompson and R. J. Hurtubise, Anal. Chim. Acta, 2006, 560,134; J. Kuijt, U. A. Th. Brinkman and C. Gooijer, Anal. Chem.,1999, 71, 1384.

15 M. E. D�ıaz-Garc�ıa, A. Fern�andez-Gonz�alez and R. Bad�ıa-La�ı~no,Appl. Spectrosc. Rev., 2007, 42, 605; I. S�anchez-Barrag�an,J. M. Costa-Fern�andez, M. Valledor, J. C. Campo and A. Sanz-Medel, TrAC-Trends Anal. Chem., 2006, 25, 958; A. S. Carretero,A. S. Castillo and A. F. Guti�errez, Crit. Rev. Anal. Chem., 2005, 35,3; R. J. Hurtubise, A. L. Thompson and S. E. Hubbard, Anal.Lett., 2005, 38, 1823; J. Kuijt, F. Ariese, U. A. Th. Brinkman andC. Gooijer, Anal. Chim. Acta, 2003, 488, 135.

16 Y. He, H. F. Wang and X. P. Yan, Anal. Chem., 2008, 80, 3832;H. F. Wang, Y. He, T. R. Ji and X. P. Yan, Anal. Chem., 2009, 81,1615.

17 X. Shu, Y. Chen, H. Yuan, S. Gao and D. Xiao, Anal. Chem., 2007,79, 3695; Y. Li, X. Liu, H. Yuan and D. Xiao, Biosens. Bioelectron.,2009, 24, 3706.

18 H. Yang, Y. Xiao, K. Liu and Q. Feng, J. Am. Ceram. Soc., 2008, 91,1591.

19 S. Chakrabarti, D. Ganguli and S. Chaudhuri, J. Phys. D: Appl.Phys., 2003, 36, 146.

20 H. L. Xia and F. Q. Tang, J. Phys. Chem. B, 2003, 107, 9175.21 Z. P. Fu, B. F. Yang, L. Li, W. W. Dong, C. Jia and W. Wu, J. Phys.:

Condens. Matter, 2003, 15, 2867.22 Q. Jiang, Z. Y. Wu, Y. M. Wang, Y. Cao, C. F. Zhou and J. H. Zhu,

J. Mater. Chem., 2006, 16, 1536.23 L. Fern�andez, N. Garro, J. E. Haskouri, M. P�erez-Cabero, J. �Alvarez-

Rodr�ıguez, J. Latorre, C. Guillem, A. Beltr�an, D. Beltr�an andP. Amor�os, Nanotechnology, 2008, 19, 1.

24 W. H. Green, K. P. Le, J. Grey, T. T. Au and M. J. Sailor, Science,1997, 276, 1826.

25 M. Anpo and Y. Kubokawa, J. Phys. Chem., 1984, 88, 5556.26 Z. Yan, Y. Ma, D. Wang, J. Wang, Z. Gao and T. Song, J. Phys.

Chem. C, 2008, 112, 9219.

This journal is ª The Royal Society of Chemistry 2010

Publ

ishe

d on

29

July

201

0. D

ownl

oade

d by

Uni

vers

ity o

f M

ichi

gan

Lib

rary

on

21/1

0/20

14 2

1:37

:52.

View Article Online

27 E. De la Rosa, S. Sep�ulveda-Guzman, B. Reeja-Jayan, A. Torres,P. Salas, N. Elizondo and M. J. Yacaman, J. Phys. Chem. C, 2007,111, 8489.

28 L. Irimpan, V. P. N. Nampoori and P. Radharishnan, J. Appl. Phys.,2007, 102, 063524(1).

29 X. M. Zhang, J. H. Zhang, X. G. Ren and X. J. Wang, J. Solid StateChem., 2008, 181, 393.

This journal is ª The Royal Society of Chemistry 2010

30 Y. Li, Y. Yang, S. Fu, X. Yi, L. Wang and H. Chen, J. Phys. Chem. C,2008, 112, 18616.

31 B. C. Cheng and Z. G. Wang, Adv. Funct. Mater., 2005, 15,1883.

32 Y. Jin, H. Wu, Y. Tian, L. H. Chen, J. J. Cheng and S. P. Bi, Anal.Chem., 2007, 79, 7176; A. Afkhami and L. Khalafi, Microchim.Acta, 2005, 150, 43.

Analyst, 2010, 135, 2386–2393 | 2393