7/25/2019 2001_0310

1/5

ournal of C atalysis 203,8286 (2001)

o i:10.1006/jcat .2001.3316, ava ila ble o nline a t htt p://ww w.ide al ibra ry.com o n

Morphology of a TiO2Photocatalyst (Degussa, P-25) Consistingof Anatase and Rutile Crystalline Phases

Teruhisa O hno, Koji Sa rukawa , Ko jiro Tokieda, and Michio Ma tsumura1

Research C enter for Solar E nergy Chemi stry, O saka U niv ersity, 1-3 Machik aneyama, Toyonak a, Osaka 560-8531, Japan

R eceived D ecembe r 8, 2000; revised Ma y 25, 2001; accepted June 23, 2001

The TiO2 powder (Degussa, P-25), which is a standard material

n the field of photocatalytic reactions, contains anatase and rutile

hases in a ratio of about 3 : 1. Transmission electron microscopy

howed that the anatase and rutile particles separately form their

gglomerates. The average sizes of the anatase and rutile elemen-

ary particles are 85 and 25 nm, respectively. Diffuse reflectancepectra of the TiO2 powder were successfully traced by physically

mixing pure anatase and rutile particles in a ratio of 3 : 1. B y the

HF treatment of the TiO2 powder, pure rutile particles were iso-

ated. All these results indicate that the rutile phase does not exist

s an overlayer on the surface of anatase particles, but it exists

eparately from anatase particles. We also found that photocat-

lytic oxidation of naphthalene is inefficient on pure anatase and

utile powders. However, the reaction is very efficient on the P-25

owder, as well as on a mixture of pure anatase and rutile particles.

Under the conditions of the photocatalytic reactions, the anatase

nd rutile agglomerates are considered to be decomposed, and the

natase and rutile particles are in contact, leading to a synergy

ffect. c 2001 Academic Press

KeyWords:photocatalyst; titanium dioxide; anatase; rutile; mor-

hology; TEM.

INTRODUCTION

Photocatalyticreactions on semiconductor powdershave

ttra cted much att ention because o f their applicability to

he treatment of a variety of pollutants and wastes (111)

n d ut ilizat ion o f sola r energy (1218). To a chieve high ef-

ciency of the reactions, many kinds of TiO 2powders haveeen examined. From the results, it has been concluded tha t

na tase particles with a large surface area a re efficient for

he decomposition of pollutants in air and water (10, 11).

n these reactions, the large surface area is especially im-

portant, because the concentrations of the pollutants are

usually very low. On the other hand, for splitting water,

which is an importa nt reaction t o convert light energy into

hemical energy, rutile particles with a small surface area

1 To who m correspo ndence should be ad dressed. Fax: 81-6-6850-6699.

-mail: matsu@chem.es.osaka-u.ac.jp.

are effecient (1618). I n this case, a ba nd bending sho

be developed in each particle to oxidize water, and hen

large particles are a dvanta geous.

In the field of TiO 2-photoca ta lyzed reactions, the pow

called P-25 (D egussa) has been a standa rd mat erial, wh

has a relatively large surface area (49 m 2 g1). Intere

ingly, the P -25 powder consists of a nat ase and rutile phaBecause the P-25 powder shows high activity for ma

kinds of photocat alytic reactions, it ha s been used in ma

studies. I n mo st studies, how ever, the effects of the mix

phases and the morphology have not been discussed

relation to the activity. The first extensive work to cor

late the photocata lytic activity and the structure of P-25 w

carried out by Bickley et al.(19). They obtained transm

sion electron microscopic (TEM) images and diffuse

flectance spectra of the P-25 powder, and they repor

that the surface of ana tase particles is transformed to

rutile structure. Afterward, based on precise TEM obsvation, D atye et al. (20) concluded that anatase and ru

single crysta lline part icles exist separa tely in the P-25 po

der. How ever, they did not differentiate the morphology

the a nat ase a nd rutile particles. U nder the circumstanc

clarification of the morphology of the mixed phases is s

of great importance to deepen the understanding of

activity of the TiO 2photoca ta lysts.

We have reported that pure rutile particles are eas

isolated from the P-25 powder by treatment with HF

lution, because the anatase phase dissolves rapidly in t

solution (21). B y comparing the activities of the origi

P-25 powder a nd the isolated rutile particles, we a lso fo u

that the presence of both ana tase a nd rutile phases is

portant for some photocatalytic reactions where oxyg

is used as the electron acceptor (21). To clarify the s

ergy between the anatase and rutile phases, we stud

the morphology of the anatase and rutile phases in P-

powder by measuring the TEM images and the diffu

reflectance spectra of the particles. From the results,

propose a model that is different from that reported

Bickley et al. (19). Our model is similar to that of Da

et al.(20) but d ifferent in the understanding o f the syne

effect.

021 9517/01 $35 00

82

7/25/2019 2001_0310

2/5

MOR PH OLO G Y O F A TiO2PHOTOCATALYST

EXPERIMENTAL

We used titanium dioxide (Degussa, P-25) powder ob-

a ined from Nippon A erosil Co. L td. For comparison, vari-

us kinds of TiO 2powders having ana ta se and rutile parti-

les were used; these were obta ined from Ishihara C o. Ltd.

ST-01, ST-21, P T-101) and Toho Tita nium C o. Ltd . (NS-51).

ontents of the anatase phase and relative surface areas of

hese powders are as followsP-25: 75%, 49.2 m2/g; ST-1: 100%, 192.5 m2/g; ST-21: 100%, 56.1 m2/g; P T-101: 0%,

5 m2/g; N S-51: 0%, 6.5 m2/g. In some e xperiments, some o f

he pure ana ta se powders and the pure rutile pow ders were

mixed in a rat io of 3: 1, which is the same as that o f the P-25

owder. For the mixing, the TiO 2pow ders were dispersed

n water, sonicated for 30 min, filtrated, and dried under

educed pressure successively at 50C for 1 h and a t 100C

or 2 h.

The morphologies of the TiO 2 powders were examined

sing a Hitachi H-800 TEM and a Hitachi S-5000 scan-

ing electro n microscope. The diffuse reflecta nce spectra o fhe pow ders were o bta ined using a Shimadzu U V-2500PC

pectrophotometer equipped with an integrating sphere.

The contents of ana tase a nd rutile phases in t he powders

were det ermined using an X -ray diffraction meter (P hilips

XP ert-MR D ). The surface a reas of the powders w ere

FIG. 1. TEM pho togra phs and electron diffra ction patterns of the P-25 powder in the regions of (a) the anata se phase and (b) the rutile phase

measured using a surface area analayzer (Micromeriti

FlowSorb II 2300).

Photocatalytic reactions were carried out at room te

perature in P yrex glass tubes ( = 15 mm), w hich cont

TiO 2pa rticles (0.015 g), naphtha lene (0.1 g), and a mixtu

of acetonitrile and wa ter (4.0g). D etails of the experimen

conditions were reported in our previous papers (22, 23

RESULTS AND DISCUSSION

TEM images of the P-25 powder clearly indicate t

the a nat ase a nd rutile particles exist separa tely by formi

their agglomerates, as shown in Fig. 1. The crystal stru

tures were assigned from the electron diffraction pa tter

From these images, the average diameters of the anata

and rutile particles are estimated to be a bout 25a nd 85 n

respectively. B y careful observation, w e found amo rpho

particles in addition to the anatase and rutile particles,

shown in Fig. 2. The content of the amorphous partic

is very low: about 1.0%. It should be emphasized that images demonstrating tha t the crysta l structure is partia

transformed in a particle were obtained, as has been

ported by D atye et al.(20).

The P -25 powder is manufa ctured by the A erosil proce

by which titanium tetrachloride is subjected to hydroly

7/25/2019 2001_0310

3/5

4 O H N O E T A L .

FIG. 2. TEM photograph of the a morphous phase included in the

-25 pow der.

n the vapor phase at an elevated temperature (19, 20).

D uring this process, the growth of particles is q uenched

when the particles (or the agglomerates) reach a certain

ize, since the grown particles (or the agglomerates) are

emoved from the rea ction zone because of their increased

weight. The TEM pictures of the P-25 powder shown in

Fig. 1 suggest that the anatase and rutile particles grow

n different nuclei during the process. In contrast to our

bservation, Bickley et al. (19) reported that ana tase a nd

utile phases in the P-25 powder often appeared in close

proximity. On the basis of their results, they suggested

hat the rutile phase forms an overlayer on the surface

f anatase particles. This model might be reasonable if

he phase transformation occurred from the surface of

he anatase particles during the manufacturing process.

H ow ever, our TE M images are strongly against the model.

The difference in TEM images obtained by Bickley et al.nd us is probably caused by the different sample prepa-

ation conditions. We sprinkled directly the particles on

arbon films, which were supported Cu meshes. On the

ther ha nd, B ickley et al. placed the particles on the

ubstrates from the suspension of butanol. Their method

s not a dequate to o bserve the morphology of the powder,

ecause the ana ta se and rutile particles mix, once they a re

uspended in solution, as discussed later.

In a ddition to the TE M images, B ickleyet al.argued that

he rutile phase is generated as a result of transformation

t the surface of ana ta se particles on the basis of the diffuseeflectance spectrum of the P-25 powder. They observed

that the diffuse reflectance spectrum of the P-25 pow

was different from that of a powder prepared by phy

cally mixing anatase and rutile particles. Based on this

sult, they proposed that morphology of the P-25 pow

is different from that of the physically mixed powd

and the spectrum of the P-25 powder was attributed

the part ly transformed crystalline structure. H ow ever, th

might have erroneously selected the anatase and ru

powders for preparing the mixed powder. In particul

systems, the scattering of light is strongly dependent

the pa rticle size. Hence, we selected t he pure ana ta se (

21, average diameter about 30 nm) and rutile (PT-1

average diameter about 75 nm) particles, the diamet

of which are similar to those of anatase and rutile pa

cles included in the P-25 powder. We mixed them in

ratio of 3: 1, w hich is identical to the ratio of the co

ponents in the P-25 powder. The detailed procedure

prepara tion o f t he mixed TiO 2 powder is described in

Experimental section. The diffuse reflectance spectra

P-25 and the mixed powder show the onset of the refltion in the region between t he onsets of pure ana tase a

rutile powders, as shown in Fig. 3. Importantly, the sp

trum of the mixed powder wa s in good agreement with t

of t he P -25 pow der. Therefore, in contrast to the report

Bickley et al., the diffuse reflectance spectra did not g

evidence supporting that partly transformed phases ex

on a particle of the P -25 powd er.

We have found that pure rutile particles are easily i

lated from the P -25 powder by trea tment w ith 10% H F

lution for 24 h (21). The diffuse refl ecta nce spectrum of

isolated pow der is shown b y a bold line in Fig. 3. The onwa velength of t he ab sorption (i.e., the band gap) is in go

agre ement w ith tha t of the pure rutile powd er (P T-101)

should be noted t hat t he reflectance of t he H F-treat ed P

FIG. 3. D iffuse reflectance spectra of pure ana tase powd er (ST-

pure rutile powde r (P T-101), the mixture of the pure a nata se and ru

powders (3 : 1), and the P-25 powder. The spectrum for the P-25 powafter the treatment with HF solution is shown by the bold line.

7/25/2019 2001_0310

4/5

MOR PH OLO G Y O F A TiO2PHOTOCATALYST

FIG. 4. TE M photograph of the rutile particlesisolated from the P-25

owder.

owder at a wa velength longer than 400 nm is lower tha n

hose of the pow ders, a s shown in Fig. 3. C onsistent with

he spectrum, the isolated rutile powder is grayish. Befo re

he HF treatment, however, the reflectance of the P-25

owder is high. This is probably because of the strong light

cattering by the small anatase particles, whose content is

5%. It is known that TiO 2 powders show absorption in

he wa velength region longer than 400 nm, if they are redu-

ed at elevated temperature and the absorption of these

owders is at tributed to the Ti3+ ions (24). He nce, the rutile

articles in the P-25 powder are considered to contain

he Ti3+ ions, which form the electron donors. Accordingly, is considered that a fa irly large band bending is generated

n the rutile particles despite their rather small particle size

85 nm). This is probab ly a n importa nt reason fo r the high

eactivity o f the P-25 pow der for many kinds of reactions.

Figure 4 shows the TEM image of the rutile particles

solated from the P-25 powder by the treatment with 10%

HF solution for 24 h. The structure is almost the same as

hat observed in the rutile particles observed in the P-25

owder, w hich is shown in Fig. 1b. In contra st, the ana tase

articles, as shown in Fig. 1a, completely disappeared after

he trea tment w ith 10% H F solution. The disappeara nce ofhe ana ta se particles also contradicts the argument that the

utile phase exists as a n overlayer on the surface of a nat ase

articles.

We found a clear synergy effect between the anatase

nd rutile phases for the photocatalytic oxidation of

aphthalene, a s shown in Table 1. The main product is

-formylcinnamaldehyde (25), which was erroneously as-

igned to 1,8-dihydro xyna phtha lene in our previous papers

22, 23). As seen in Table 1, the mixt ure of ST-01 (a na ta se)

nd NS-51 (rutile) showed the activity as high as that of

-25, w herea s the act ivity o f pure ST-01 an d N S-51 pow -ers is low.

TABLE 1

Photocatalytic Oxidation of Naphthalene on TiO2Powders

TiO 2pow der A mount of 2-formylcinna na ldehydea/mo l

P -25 13.4

ST-01 (a na ta se) 1.4

NS-51 (rutile) 4.3

ST-01 : NS-51 (3 : 1) 11.5

a Product was quantitatively analyzed by H PLC after photoirra-

diation for 1 h.

The synergy effect can b e expected, if rutile and a nat a

particlesareincontact.Figure5showsthescanningelectr

microscopic image o f t he mixture o f pure a na ta se (ST-0

a nd pure rutile (NS-51) pow ders, which were mixed by so

ication in w ater. B efore the mixing, the pow ders form

glomerates, as the anatase and rutile phases in the P-

powder. Since the sizes of the anatase and rutile partic

in these two kinds of powders are largely different, ocan easily see that t he ana tase particles are loaded o n t

rutile particles as shown in Fig. 5. In the case of the P-

powder also, the anatase and rutile particles are mixed

a similar fashion after the same treatment. However, su

a clear mixed image as shown in Fig. 5 was not obtain

for P -25, because the difference in the sizes of ana ta se a

rutile particle is not large enough. I n the photo cata lytic

a ctions, the solution is sonica ted t o suspend the TiO 2pa

cles before photoirradiation, and the anatase particles a

FIG. 5. Scanning electron microscopic photograph of a mixture

ST-01 (pure ana tase) and NS-51 (pure rutile) pow ders. The elemen

particle sizes of the powders are 7.0 and 220 nm, respectively. From

difference in the pa rticle sizes, the small dot s deposited on t he large ruparticles are assigned to the agglomerates of fine anatase particles.

7/25/2019 2001_0310

5/5

6 O H N O E T A L .

probably loaded on the rutile particles. This is why synergy

ffect appearsfo r the mixture of ana tase and rutile particles,

which make separated agglomerate before sonication.

CONCLUSIONS

We ha ve found tha t the ana ta se and rutile particles exist

epara tely in the P-25 (D egussa) powder. B y the TE M ob -

ervat ion, the agglomerates made of either ana ta se or rutileparticles are clearly observed. However, under the practi-

al o perational conditions of the photocatalytic reactions,

he agglomerates are expected to be decomposed, and the

natase particles and rutile particles are in contact. This

mixed structure must be the key to the high a ctivity o f the

-25 pow der.

R EF ER ENCES

1. Schwitzgebel, J., Ekerdt, J. G ., G erischer, H., and H eller, A.,J. Phys.

Chem. 99, 5633 (1995).

2. Z ang, L., La nge, C., Abraha m, I., Storck, S., Maier, W. F., and K isch,H ., J. Phys. Chem. B102, 10765 (1998).

3. Ohnishi, H., Matsumura, M., Tsubomura, H., a nd I wasaki, M., I n d .

E ng. Ch em. Res. 28, 719 (1989).

4. Mat sumura, M., Furukawa , S., Saho, Y., and Tsubomura, H .,J. Phys.

Chem. 81, 1327 (1985).

5. Jackson, N. B., Wang, C. M., Luo, Z., Schwitzgebel, J., Ekerdt, J.

G. , B rock, J . R . , a nd Heller , A . , J. El ectrochem. Soc. 138, 3660

(1991).

6. Ho ffmann, M. R ., Martin, S. T., Choi, W., and B ahnemann, D. W.,

Chem. Rev. 95, 69 (1995).

7. Fujishima , A . , Ca i , R . X. , Ot suki, J ., Ha s himot o, K. , I t oh,

Yama shita, T., and K ubota , Y.,El ectrochim. A cta18, 153 (1993).

8. D Oliveira, J. C ., Minero, C., P elizzetti, E ., and Pichat, P., J. Ph

chem. Photobi ol. A 72, 261 (1993).

9. Yamashita, H ., Miwa, H., Harada, M., Ichihashi, Y., Anpo,

J. Phys. Chem. B102, 10707 (1998).

10. R a o, M . V. , R a jeshw a r, K. , Vernerker, V. R . , a nd Kubow

J. Phys. Chem. 84, 1987 (1980).

11. Nishimoto, S., Oht ani, B., Ka jiwara, H ., and Ka giya, T.,J. Chem. S

Faraday Tr ans. 181, 61 (1985).12. Sato, S., and White, J. M.,J. Catal. 69, 128 (1981).

13. Sayama, K., and Arakawa, H., J. Phys. Chem. 97, 531 (1993).

14. B ard, A. J., and Fox, M. A.,A cc. Ch em. Res. 28, 141 (1995).

15. Takat a, T., Furumi, Y., Shinohara, K ., Tana ka, A., H ara, M., Kon

J. N., and D omen, K., Chem. M ater. 9, 1063 (1997).

16. Ohno, T., Ha ga, D ., Fujihara, K., K aizaki, K., and Matsumura,

J. Phys. Chem. B101, 6415 (1997); E rra ta ,101, 10605 (1997).

17. Fujihara , K., O hno, T., and Matsumura, M., J. Chem. Soc., Fara

Trans. 94, 3705 (1998).

18. O hno, T., Fujihara, K ., Sarukawa , K., Tanigaw a, F., and M atsum

M. , Z . Phys. Chem. 213, 165 (1999).

19. B ickley, R. I., G onzalez-Ca rreno, T., Lees, J. S., Palmisano, L .,

Tilley, R. D., J. Soli d State Chem. 92, 178 (1991).

20. D atye, A. K., Riegel, G ., Bolton, J. R., Huang, M., and P rairie, MJ. Soli d State Chem. 115, 236 (1995).

21. Ohno, T., Sarukawa, K., a nd M atsumura, M., J. Phys. Chem. B

2417 (2001).

22. Jia, J., Ohno, T., Masaki, Y., and Matsumura, M., Chem. L ett.

(1999).

23. Jia , J., Ohno, T., and Ma tsumura, M.,Chem. L ett.908 (2000).

24. R othenberger, G ., Moser, J., G ratzel, M., Serpone, N., and Shar

D. K., J. Am . Chem. Soc.107, 8054 (1985).

25. Soana, F., Sturini, M., Cermenati, L., and Albini, A., J. Chem. S

Perkin Trans. 24, 699 (2000).