MOO - American Chemical Society

---

1484

J . Phys. Chem. 1989, 93, 1484-1490

Surface Active Sites of V,O,-MOO,

Catalysts

Atsushi Satsuma,* Atsushi Hattori, Koichi Mizutani, Akio Furuta,+ Akira Miyamoto,f Tadashi Hattori, and Yuichi Murakami Department of Synthetic Chemistry, Faculty of Engineering, Nagoya University, Furo-cho, Chikusa- ku, Nagoya 464, Japan (Received: April 25, 1988)

The promoting effect of Moo3 on the surface active sites of V2O5 catalyst was investigated by using various methods, and the catalytic activity of V205-MOO, catalysts was discussed in the light of the surface active sites and the bulk structure. The surface concentration of the redox sites, measured by the NARP (NO-NH, rectangular pulse) technique, was increased by the addition of Moo3 to V 2 0 5 both on the solid solution and on the intermediate compounds, Le., Mo4V602, and Mo6V90G On the solid solution, the increase in the surface redox sites is attributed to the formation of the surface V=O species on various crystal planes of vanadium oxide containing solved MOO,, although V2O5 crystal has the surface V=O species only on the (010) plane. The increase in the surface V=O species on solid solution may cause the high activity of V205-MOO, catalysts on various reactions, such as benzene oxidation. On the intermediate compounds, it was proved by NARP and SIMS that the surface Mo=O species act as the redox sites for the reduction of NO with NH3 in a manner similar to that of the surface V=O species on V 2 0 5 catalysts, although Moo3 itself is inactive at the temperature range examined. The activation of the surface MFO species was attributed to the atomic mixing of the surface vanadium and molybdenum ions on the exposed crystal plane of the intermediate compounds.

Introduction

Experimental Section

The addition of MOO, improves the activity and selectivity of V2O5 catalysts for oxidation of hydrocarbon^.'-^ Actually, V205-Mo03catalysts have been widely used in benzene oxidation for the industrial production of maleic anhydride.6 The activity and selectivity of V205-Mo03 catalysts varies significantly with MOO content, and for example, in the benzene oxidation, the maximum selectivity has been obtained at around 30 mol %

V205-Mo03 catalysts were prepared by mixing oxalic acid solution of NH4V03with (NH4)6Mq024-4H20 followed by drying and subsequent calcination in flowing O2at 773 K for 3 h. The atomic ratios of Mo/(V Mo) were 0.1, 0.3, 0.5, 0.7, and 0.9. V2Os and Moo3 were prepared by the thermal decomposition of N H 4 V 0 3and ( N H , ) , M O ~ O ~ ~ . ~ H in, O flowing o2at 773 K for 3 h, respectively. The catalysts were then pressed and sieved in the range of 28-48 mesh. X-ray diffraction diagrams of the catalysts were obtained with a Rigaku GF 2035 X-ray diffractometer with Cu target. IR spectra of the catalysts were recorded on a Jasco FT/IR-3 Fourier transform infrared spectrometer with KBr as a diluent. ESR absorption measurements were made at X-band on JEOL ME

MOO^.^

Since the crystal phase and the structure also vary significantly with MOO, content, it is expected that the catalytic activity and selectivity of V205-Mo03 catalysts are closely related to the change of the crystal phase. Many investigations have been devoted to the identification of an active phase for the selective oxidation of benzene. Solid solution of molybdenum in V2O5 is expected to be one of the effective phases in selective oxidation of benzene and other hydrocarbons, since the maximum selectivity is frequently obtained at the limit of existence of the solid solut i o ~ ~ .Tarama ~ - ~ et al.397 suggested that the formation of V205-Mo03 solid solution weakens V=O bond, which is widely accepted as the active site for redox process.* Dyrec et aL9 reported the role of vanadium and molybdenum ions in solid solution as the redox center in the oxidation process. Intermediate compounds are also expected to be the active phase. Munch et a1.I0 suggested Mo6V90mand Mo4V6025, and Kuznetsova et al." suggested VMo3011.Najbar et al.I2 also extensively investigated the intermediate compounds in the subsurface layer of the catalyst and their phase transitions during benzene oxidation. In spite of much literature about the bulk structure of V20 5 - M d 3 catalysts, little is known about the catalyst surface which has a direct effect on the catalysis. A better understanding of the promotion effect of MOO, on V2O5 catalysts may be derived from the investigation of the surface active sites, i.e., the surface V=O species and the acid sites. We have reported that P 2 0 ~ ' ~ and wO3l4have a large effect on the surface active sites. In this work, we have examined the active sites of the V2OS-MoO3 catalysts by using the NARP (NO-NH, rectangular pulse) technique and so on, and the promotion effect of MOO, on the surface active sites has been discussed in light of the bulk structures such as solid solution and intermediate compounds. The activity of V205-Mo03catalysts for some oxidation reactions also has been discussed on the basis of the surface active sites.

'

Kinu-ura Research Department, JGC Corporation, Sunosaki-cho, Handa, Aichi 475, Japan. *Present address: Department of Hydrocarbon Chemistry, Faculty of Engineering, Kyoto University, Sakyo-ku, Kyoto 606, Japan.

0022-3654/89/2093-l484$01.50/0

+

(1) (1) (a) Tarama, K.; Teranishi, S . ; Yasui, A. Kogyo Kagaku Zassi 1957, 60, 1222. (b) Ioffe, I. I.; Lyubarskii, A. G. Kine?. Katal. 1962, 3, 261. (c) Bhattachargya, S. K.; Venkataraman, N. V. J . Appl. Chem. 1958,8,728. (d) Cole, D. J.; Cullis, C. F.; Hucknall, D. J. J . Chem. Soc., Faraday Trans 1 1976, 72, 2744. (2) Ioffe, I. I.; Ezhkova, Z. I.; Lyubarskii, A. G. Kine?. Katal. 1962, 3, 165.

(3) Tarama, K.; Teranishi, S.; Yoshida, S.; Tamura, N. Proc. Int. Congr. Catal. 1965, 3, 282. (4) Ai, M. Bull. Chem. SOC.Jpn. 1971, 44, 761. (5) Ono, T.; Kubokawa. Y. Bull. Chem. SOC.J m 1982, 55, 1748 (6) U S . Patent 3,005,831, 1961. Jpn. Kokai '135, 1963. U S . Patent 3,221,671, 1965. US.Patent 3,417,108, 1968. (7) Tarama, K.; Yoshida, S.; Ishida, S.; Kakioka, H. Bull. Chem. SOC.Jpn. 1968, 41, 2840. (8) (a) Hirota, K.; Kera, Y.; Teratani, S . J . Phys. Chem. 1968, 72, 3133. (b) Nakamura, M.; Kawai, K.; Fujiwara, Y. J . Caral. 1974, 34, 345. (c) Cole, D. J.; Cullis, C. F.; Hucknall, D. J. J . Chem. Soc., Faraday Trans. I 1976, 72, 2185. (d) Akimoto, M.; Usami, M.; Echigoya, E. Bull. Chem. SOC.Jpn. 1978, 51, 2195. (e) Agarwal, D. C.; Nigam, P. C.; Srivastava, R. D. J . Catal. 1978, 55, I . (0 Bond, G. C.; Sarkany, A. J.; Parfitt, G. D. J . Catal. 1979, 57, 176. (8) Miyamoto, A,; Yamazaki, Y.; Inomata, M.; Murakami, Y. Chem. Letr. 1978, 1355; J . Phys. Chem. 1981, 85, 2366. Inomata, M.; Miyamoto, A,; Murakami, Y. Ibid. 1981,85, 2372. (h) Nakagawa, Y.; Ono, T.; Miyata, H.; Kubokawa, Y. J . Chem. Soc., Faraday Trans. 1 1983, 79, 2929. (i) Jansen, F. J. J. G.; Kerhoff, F. M. G.; Bosch, H.; Ross, J. R. H . J . Phys. Chem. 1987, 91, 5921. (9) Dyrek, K.; Labanowska, M. J . Catal. 1983, 81, 46; Ibid. 1985, 96, 32; Appl. Catal. 1986, 23, 63. (10) Munch, R. H.; Pierron, E. D. J . Catal. 1964, 3, 406. ( 1 1) Kuzenetsova, T. G.; Boreskov, G.K.; Andrushkevich, T. V.; Plyasova, L. M.; Maksinov, N. G.; Olenkova, I. P. React. Kine?. Catal. Lett. 1979, 12, 531. (12) Najbar, M. Proc. In?. Congr. Catal. 1984, 5, 323. Najbar, M.; Stadnicka, K. J . Chem. SOC.,Faraday Trans. I 1983, 79, 27. Najbar, M.; BielaAski, A,; Camra, J,; Bielaiiska, E. Int. Symp., Prap. Catal. 1986, 4 , BS. Najbar, M. J . Chem. Soc., Faraday Trans. 1 1986, 82, 1673. (13) Satsuma, A,; Hattori, A,; Furuta, A,; Miyamoto, A,; Hattori, T.; Murakami, Y . J . Phys. Chem. 1988, 92, 2275. (14). Satsuma, A,; Hattori, A,; Mizutani, K.; Furuta, A,; Miyamoto, A,; Hattori, T.; Murakami, Y . J . Phys. Chem. 1988, 92, 6052.

0 1989 American Chemical Society

The Journal of Physical Chemistry, Vol. 93, No. 4, 1989 1485

Surface Active Sites of V205-MOO, Catalysts

(CRl-1)

1200 20

I

I

30

P o

40

50

1

60

I v205

1000

1

1

\, 1022

800

1

600

1

1

400

1

~

1

I829

Mo/( V+Mo)

i[ =0,9

=

0.1

=

0.3

=

0.5

=

0.7

=

0,9

Figure 1. XRD diagrams of V205-Mo03 catalysts with various compositions: (0)v 2 0 5 ; (A) M o o 3 ; ( 0 ) M o 6 V @ ~ ;(V) M o ~ V ~ ~ ~ ~ .

1 X spectrometer at room temperature. X-ray photoelectron spectra (XPS) of the catalysts were measured with a Shimadzu ESCA-750 with Mg KP radiation (1253.6 eV). SIMS spectra of the catalysts were recorded with VG ESCA LAB-5. In order to avoid charging of the samples, a charge neutralizer was used under the optimum conditions for each sample. In situ infrared spectra of adsorbed N H 3 were measured with Jasco FT/IR-3 spectrometers having an infrared cell connected to a conventional high-vacuum system. TPD spectra of adsorbed NH3 were measured with a conventional TPD apparatus similar to that described by Amenomiya et al.15 The numbers of surface V=O species on the catalysts were determined by using the NARP (NO-NH3 rectangular pulse) technique.I6 All the experiments were carried out under the conditions described p r e v i ~ u s l y . ' ~ , ' ~

Results X-ray Diffraction. Figure 1 shows XRD diagrams of VzO5MOO, catalysts with various compositions. At Mo/(V + Mo) = 0.1, V2O5 (ASTM 9-387) phase was the most abundant, though less intensive diffraction lines assignable to M o ~ (ASTM V ~ ~ 19-812) were also observed. The lattice distance of V2O5 along the c axis decreased from 4.38 to 4.30 A, indicating the formation of solid s o l ~ t i o n . ~ *At ' ~ Mo/(V + Mo) = 0.3, diffraction lines assignable to V2OS became weak, and those of Mo4V6025and MOgV9040 (ASTM 19-813) became intense. The line observed at 28 = 37.8' ( d = 2.380 A) could not be assigned. At Mo/(V + Mo) = 0.5, lines assignable to intermediate compounds, Le., M o ~ and V ~ ~ ~ were ~ the most abundant, indicating the catalyst was mainly composed of intermediate compound phases. Diffraction lines assignable to MOO, (ASTM 5-0508) became significant at Mo/(V + Mo) = 0.7, and only Moo3 phase was observed at Mo/(V Mo) = 0.9. ZR Spectra. IR spectra of V205-MO03 catalysts are shown in Figure 2. V2O5 exhibited a band at 1022 cm-l which is assignable to the V-0 stretching v i b r a t i ~ n . ~When ~ ~ ~ 'MOO, ~

+

(15) CvetanoviE, R. J.; Amenomiya, Y. Ado. Coral. 1967, 17, 103; Cafal. Rev. 1972, 6 . 21. (16) Inomata, M.; Mori, K.; Miyamoto, A,; Ui, T.; Murakami, Y. J . Phys. Chem. 1983, 87, 754. Inomata, M.; Mori, K.;Miyamoto, A.; Murakami, Y. Ibid. 1983, 87, 754. (17) Guidot, J.; Arnaud, Y.; Germain, J. E. Bull. SOC.Shim. Fr. 1980, 1-209.

l

1200

,

11991

1000

I879

800

I

600 l l

400 1

'iovenumber ( c m - l )

Figure 2. IR spectra of V,05-Mo03 catalysts with various compositions.

was added to VzO5, the absorption band weakened and shifted to lower wavenumber, which is in good agreement with the results obtained by Tarama et al.3+7At Mo/(V Mo) = 0.3 and OS, a new band was observed at around 960 cm-I. At Mo/(V + Mo) = 0.9 and Moo3, the M e 0 stretching bandlg was observed at 991 cm-I, instead of the V=O stretching band. V205 showed a combination band of V-0-V stretching and lattice vibrati~n~g'-'~ at 829 cm-I, and M o o 3 also showed a combination band of Mo-0-Mo stretching and lattice vibration19 at 879 cm-'. All of the V205-MOO, catalysts showed a similar absorption band between 829 and 879 cm-I. ESR Spectra. As shown in Figure 3, ESR signals were observed ~at g~ = 1.96 on the catalysts except V2O5 and MOO,. Although both V4+and Mo5+ ions give an ESR signal, the g-value indicates that the signal can be assignable to V4+ions. Hyperfine structure was observed only at Mo/(V + Mo) = 0.9, suggesting that the vanadium ions are well isolated in the catalyst. Table I shows the relative amount of V4+ ions determined by double integrations. The amount of V4+ ions was the largest at Mo) = 0.1, and it decreased with increasing MOO, Mo/(V content. XPS Spectra. Figure 4 shows the binding energies of XPS peaks of V(2p), M0(3d,/2), and M0(3d5/2). The binding energies were corrected with a reference to the C( Is) peak at 285.0 eV. As shown in Figure 4, t h e binding energy of V ions decreased b y the addition of MOO,. On the other hand, the binding energy of Mo(3d) did not vary with the composition.

+

+

(18) (a) Frederickson, L. D.; Hansen, D. M. Anal. Chem. 1963,35,818. ( b ) Miyamoto, A.; Yamazaki, Y.; Murakami, Y. Nippon Kagaku Kaishi 1977, 619. Inomata, M.; Miyamoto, A,; Murakami, Y. J. Cafal. 1980, 62, 140; Chem. Left. 1978, 799. (19) Trifilo, F.; Pasguon, I. J . Catal. 1968, 12, 412. Mitchell, P. C. H.; Trifilo, F. J. Chem. SOC.A 1970, 3183. Trifilo, F.; Notarbartolo, S.;Pasquon, L. J . Catal. 1971, 22, 324.

1486 The Journal of Physical Chemistry, Vol. 93, No. 4 , 1989

Satsuma et al.

TABLE I: Relative Amount of V4+" BET Surface Area (SBm), Number of Redox Sitesb ( L ) ,Surface Concentration of Redox Sites ( L / S B ~ ) , Number of Acid Sites (A), Surface Concentration of Acid Sites ( A / S B ~and ) , Sum of Surface Concentrations of Redox Sites and Acid Sites ( ( L A )/SBm) of V2O5-MoO3 Catalysts

+

catalyst

SBET.

amt of V4+c

v2°5

+ Mo) = 0.1 + Mo) = 0.3 + Mo) = 0.5 Mo/(V + Mo) = 0.7 Mo/(V + Mo) = 0.9 Mo/(V Mo/(V Mo/(V

Moo3

0 1.o

0.47 0.18

0.47 0.04 0

L,

LISBET,

m2/g 5.8

pmol/g 22

3.8

2.3 3.7 2.7 5.3

19 47

8.3

37

10.4

46 47

1.9

0

pmol/g 43

pmol/m2 7.4 5.2 6.5 7.2 5.8

12 24

12.7 13.7

19

8.7

31

4.5 0

40 8.4

(L+ ~/SBET, pmol/m2 11.2

AISBET,

A,

pmol/m2

13.5 19.2 20.9 14.5

8.3 4.4

3.8

4.4

+ Mo) = 0.1 was fixed

Relative intensities of V4+ calculated from double integration of ESR spectra. (The intensity of ESR spectrum at Mo/(V at 1.0.) bSee text. cArbitrary units.

~ O / ( V + ~ O =I 0 , l

=

0.3 3 , 2x103 232,O

=

0,5

t 6,3xlO

3 7 =

0,7

0.9

=

0,7

0.9 1,O

Mo/(V+Mo) i Figure 4. Binding energy of V(2p3,,), Mo(3d3/J, and M0(3d5/2) based upon C(1s) = 285.0 eV. ,-.

1.0

(

I

5,ox1o3

L 8,9x103

/

Figure 5 shows the surface Mo content determined from the intensities of V(2p312) and M 0 ( 3 d ~ , ~with ) various bulk compositions. Surface Mo contents were higher than those of bulk in V,05-Mo03 catalysts except Mo/(V Mo) = 0.9, which is in good agreement with the results obtained by Bielafiski et aLm The surface enrichment of Mo may be attributed to higher sublimation pressure of Moo3 than that of V205. SZMS Spectra. Figure 6 shows SIMS spectra of (a) fresh catalyst of Mo/(V Mo) = 0.5, (b) Mo/(V Mo) = 0.5 after the redox treatment, and (c) a mechanical mixture of V2O5 and MOO, containing 50 atom % of Moo3 after the redox treatment. The redox treatment was done as follows; the catalysts, which had at 773 K for 20 min, was reduced been preoxidized in flowing 1602 by injecting several pulses of NO-NH3 mixture at 588 K. Then, the catalyst was reoxidized with 4.73 kPa of I8O2at 573 K for 4.5 h. Spectrum a indicates secondary ions of V+, VO+, Mo+, and MOO+together with satellite peaks ascribable to VH+ and VOH+. Spectrum b is essentially the same as that of the fresh catalyst

+

+

0,5

B u l k corliDosition

/

Figure 3. ESR spectra of V205-Mo03catalysts with various composi-

tions.

0,3

0 0.1

+

(20) BielaAski. A.; Camra, J.; Najbar, M. J . Catal. 1979, 57, 326.

0

o

I

0,2

I 0.4

I

I

0.6

0,8

1.0

B u l k composition (MO/(V+MO) 1 Figure 5. Relationship between surface Mo content and bulk composi-

tion. shown in spectrum a, except for satellite peaks of oxygen-containing ions. For example, the fresh catalyst gave VO+ ion peak at m / e 67, while the redox-treated catalyst gave not only the peak at m l e 67 but also a new VO' ion peak at m l e 69. The latter peak can be assignable to slV'80+, indicating that V=O was replaced with gaseous I8O2in the redox treatment. According to the reaction mechanism of the reduction of N O with N H 3 in the presence of O2proposed by us,21the surface V=O is reduced to V-OH, and the V-OH species thus formed is reoxidized to the V=O species by gaseous 02.Thus, the VI8O+ ion peak indicates that the surface V=O species act as the redox sites. The same results were obtained for MOO+ion peak. As shown in spectrum b, the redox-treated catalyst gave a satellite peak at (21) Inomata. M.: Mivamoto. A,: Murakami. Y . J . Catal. 1980. 62. 140. Miyamoto, A,; Inomata, M.;Hattori, A ; Ui,T ; Murakami, Y . J. Mol. Catal

1982, 16, 3 1 5

The Journal of Physical Chemistry, Vol. 93, NO. 4, 1989 1487

Surface Active Sites of V20S-Mo03 Catalysts Mo'

V+

Range = 300

Mo/(V+Mo)

=

0.1

=

0.3

=

0,5

I

MoVO;

=0.7

I

50

I

60

I

70

I

80

I

I

30

100

I

110

I

I

1

I

I

I

I

I

I

140

150

160

170

180

190

200

210

220

m / e (o.m,u.)

I

120

r / e la - . u

Figure 7. S I M S spectra of V205-MOO, catalysts with various compositions in the range of m / e 140-230 (resolution = 6.5).

Figure 6. S I M S spectra of (a) Mo/(V + Mo) = 0.5 and (b) Mo/(V + Mo) = 0.5 after redox treatment and (c) mechanical mixture of V 2 0 5 and MOO, after redox treatment (resolution = 8.6).

m / e 1 18 assignable to lmMo180+,which could not be observed in the fresh catalyst (a). This result indicates that oxygen bonded to Mo can be substituted with gaseous oxygen in the redox treatment; Le., the oxygen bonded to Mo also acts as the redox site. The mechanical mixture also gave the VIEO+ion peak after the redox treatment, but not the loOMo1*O+ion peak, as shown in spectrum c. This result indicates that the surface V = O species are active even without the effect of Moo3 and that the oxygen bonded to Mo is active only under the effect of V2Os. Figure 7 shows S I M S spectra in the range of m / e 140-230. Since the ion peaks were very weak in this range of mass number, they could be observed only at a low resolution. At Mo/(V + Mo) = 0.1, V203+ion gave the highest peak, and composite ions including both vanadium and molybdenum, such as MoVO', MoV02+,and MoV03+,were also observed with small intensities. With increasing Moo3 content, the intensity of MoV02+ ion relative to V203+ion remarkably increased, and it was the highest at Mo/(V Mo) = 0.5. It should be noted that such composite ions were not observed in a mechanical mixture of V205and Moo3, though V203+ and Moo3+ions were observed. It follows that the composite ions do not arise from the recombination of fragment ions, but from the surface V U M o bond on the surface of V20s-Mo03 catalysts. At Mo/(V + Mo) = 0.9, the intensity of MoV02+ became small, suggesting that V-0-Mo bonds are less abundant on the catalyst surface containing a large amount of Moo3. IR Spectra of Adsorbed NH3. Figure 8 shows infrared spectra of N H 3 adsorbed on V~OS-MOO~ catalysts. An absorption band assignable to the bending vibration of NH4+ ionz2was observed

(

1600

1800 I

Mo/ ( V+Mo ) =

0,3

=

0.7

Moo3

I

I

1200

1400 I

l

I

2

11419

1800 I

I

1600 1

1

1400 1

1

1200 1

Wavenumber (cm-')

Figure 8. Infrared spectra of NH3 adsorbed on V205-MOO, catalysts at room temperature.

at 1413-1419 cm-l, indicating the presence of the Bransted acid sites on the surface of all V2O5-M0O3catalysts. A band assignable to coordinately held NH322was observed at 1265 cm-I, suggesting the presence of Lewis acid sites on the surface of the catalysts. However, the amount of Lewis acid sites is rather small N

H.Adu. Catal. 1976, 25,

184.

cm-')

I

y 110%

+

(22) Krozinger,

1 230

1488 The Journal of Physical Chemistry, Vol. 93, No. 4, 1989

Satsuma et al.

"2'5

-.-..-.-.-.2ol

El-

5 =0.3

30

-o--cg-o-

=0.5

A

10

0520

-

540

560

580

600

620

=0.7 Temperature / K

=0.9

Figure 11. Amount of initial sharp N2measured at various temperatures: ( 0 ) V205; (0) Mo/(V + Mo) = 0.1; ( 0 ) Mo/(V Mo) = 0.3;(A) Mo/(V + Mo) = 0.5; ( 0 )Mo/(V + Mo) = 0.7; (V) Mo/(V + Mo) = 0.9; (A) MoOj.

+

7

w- 152

I 400

I

300

I 500

I

I

600

700

Mo/(V+Mo) i n b u l k

0.1

0

0,3

0,5

0 , 7 0.3

1.0

Temperature / K

Figure 9. TPD spectra of desorbed N H 3 from V20S-Mo03 catalysts with

various composition.

,578 K o

y

u

0

0.2

0.4

0,6

0.8

m 3

Mo content ( Mo/(V+Mo) ) Figure 12. Relation of surface concentrationof active sites with surface Mo content: (0)redox sites (see text); (A)acid sites; (0) sum of the surface concentrationsof the redox sites and the acid sites; (- - -) calculated surface concentration of V ions on all planes; (-) calculated surface concentratation of V and Mo ions. Upper lines were calculated from the closest packing of V 0 6 octahedral, and lower lines were calculated from the surface concentration of V ions on V,O,(OlO) plane. Surface

3

30

90

60

Time / sec

Figure 10. Concentration profiles of N 2 produced by reaction of the rectangular pulse of NO and N H 3 with Mo/(V Mo) = 0.5 at various

temperatures.

+

relative to that of Bransted acid sites from the intensity of absorption band. Another band around 1600 cm-', which is also assignable to the coordinately held NH3, was not observed. Temperature-Programmed Desorption of NH3. Figure 9 shows TPD spectra of desorbed NH3. For the unpromoted V205,the desorption maximum was observed at 343 K, and all of NH3 was desorbed below 523 K. M o o 3 and V20,-Mo03 catalysts gave almost the same TPD profiles as that of V20s. Table I shows the number of acid sites (A) determined from the amount of the desorbed NH3 in TPD experiments and the surface concentration of acid sites (A/SBET),Le., the number of acid sites per unit surface area. The surface concentration of acid sites slightly decreased with increasing M o o 3 content.

NARP Technique. When a rectangular pulse of N O and N H 3 mixture is introduced onto V 2 0 5catalyst, N2with a concentration profile consisting of initial N2 and tailing N z is observed in the temperature range from 520 to 620 K. The number of redox sites, Le., the surface V = O species, can be determined from the amount of initial N2.16 In contrast to this, M a 3 did not give any Nz peak, which agrees well with the fact that Moo3 is active only above 620 K for the reduction of NO with NH3.23 Figure 10 shows a typical example of concentration profiles of N 2 produced in the NARP technique. The profile was very similar to that observed on the unpromoted V 2 0 5 catalyst, and the initial sharp N2.can be separated from the tailing N2 in the same manner as precisely described in the previous papers.I3J4 As shown in Figure 11, the amount of initial sharp N2 was constant and independent of the reaction temperature, indicating that the NARP technique can be applied for V205-Mo03 catalysts to determine the number of surface redox sites. Table I summarizes the number of redox sites ( L ) , i.e., the surface V=O species, and the surface concentration of the redox (23) Murakami, Y.; Hayashi, K.; Yasuda, K.; Ito, T.; Minami, T.; Miyamoto, A. Nippon Kagaku Kaishi 1977, 1 1 3 .

Surface Active Sites of V2O5-Moo3 Catalysts sites (LISBET). The surface concentration of the redox sites (L/SBET) is plotted against the surface Mo content determined by XPS in Figure 12. The surface concentration of the redox sites also increased with increasing MOO, content below Mo/(Mo V) = 0.5, but further increase in Mo content reduced the surface concentration. The figure also shows the surface concentration of the acid sites (A/SBET) and the sum of the surface redox sites and the acid sites ( ( L A)/SBm). Dotted lines in Figure 12 show the calculated value of the surface concentration of cations. The sum of the surface concentrations of both sites are between two dotted lines at Mo/(V Mo) = 0.3 and 0.5, indicating that all the surface vanadium and molybdenum ions act as the active sites: either of the redox sites or of the acid sites.

+

+

+

Discussion Structure of Catalysts. M o / ( V + Mo) = 0.1. The XRD diagram (Figure 1) shows that the catalyst is composed of a large amount of V2O5 phase and a small amount of intermediate compounds. The diffraction line of V2O5 shifted along the c axis of V2O5 crystal, indicating the formation of the solid s o l ~ t i o n . ~The *’~ IR spectrum (Figure 2) also indicated the formation of solid solution, because the V=O stretching band sifted significantly. It has been reported that the V205-Mo03 system easily forms the solid solution in the composition range below 30 mol 76 of Mo03,2,3s17 Le., below Mo/(V Mo) = 0.18. ESR spectra (Figure 3) show that the reduced vanadium ions, V4+, are formed in the bulk. The binding energy of V(2p) peak in XPS spectra (Figure 4) also shows that the surface vanadium ions are reduced. These results seem quite reasonable, because the dissolution of six-valent molybdenum ions in the matrix of five-valent vanadium oxide will result in the formation of four-valent vanadium ions due to the valence control. Thus, the bulk of the catalyst of Mo/(V + Mo) = 0.1 mainly consists of V205 solid solution containing dissolved molybdenum ions and a significant amount of V4+ ions. Mol( V Mo) = 0.3 and 0.5. Catalysts of Mo/(V + Mo) = 0.3 and 0.5 are mainly composed of the intermediate compounds, i.e., Mo6V90Mand MOqV6025 (Figure 1). The IR absorption band (Figure 2) at -960 cm-l may be assignable to these intermediate compounds. Composite secondary ions in SIMS spectra (Figure 7) suggest that the V-0-Mo bonds formed on the surface of the catalysts, which is in good agreement with the fact that the catalyst bulk consists of the intermediate compounds. Thus, the structural feature of these catalysts is the atomic-scale mixing of vanadium and molybdenum ions in both the surface and the bulk. Mol( V + Mo) = 0.7 and 0.9. Diffraction lines assignable to MOO, phase became predominant (Figure l ) , and IR spectra (Figure 2) of Mo/(V + Mo) = 0.9 agreed well with that of MOO,. Thus, these catalysts are composed of a large amount of MOO, phase. With increasing MOO, phase in the bulk, the composite ions in SIMS spectra such as MoV02+ decreased (Figure 7), indicating that the V-O-Mo bonds on the surface of the catalyst decreased by the addition of a large amount of MOO,. Acid Sites. The acidic properties of the catalysts were characterized by the TPD of NH, and the IR spectra of adsorbed NH,. The amount of acid sites determined from the TPD was slightly higher on VZO5 than that on MOO,, and it decreased gradually with the increase in the surface Mo content (Figure 12). The TPD spectrum of NH3, which gives a measure of the distribution of acid strength, on V2O5 was similar to that on MOO, (Figure 9). The IR spectra of adsorbed NH, (Figure 8) indicate that the acid sites on V205-MOO, catalysts are composed of a large amount of Bransted acid sites and a small amount of Lewis acid sites. All of these results show only a little change in the type, amount, and strength of acid sites. However, this does not result from the catalyst being the mixture of large particles of V205 and MOO,. XRD diagrams in Figure 1 shows the formation of compounds, Le., Mo6V90Mand M o ~ V The ~ ~composite ~ ~ . ions in SIMS shown in Figure 7 strongly indicate that vanadium and molybdenum ions are well mixed in atomic scale on the surface. Above-mentioned results may indicate that the acidic properties of the surface vanadium and molybdenum ions are not modified significantly even if they are mixed in the atomic scale.

+

+

The Journal of Physical Chemistry, Vol. 93, No. 4, 1989 1489 Redox Sites. Mol( V + Mo) = 0.1. A small amount of added MOO, increases the surface V = O species on the catalyst (Figure 12), the bulk of which consists of the V2O5 solid solution containing Mo ions. Similar results were obtained in the case of the V205-P205 catalyst^'^ and V205-W03 catalyst^,'^ and the increase in the surface concentration of the V=O species had been attributed to the formation of lower valent vanadium ions in the bulk. It has been proved by the crystallographic study of Anand the experimental studies by our groupZ5that lower valent vanadium oxides have a higher concentration of the surface V = O species than that on V2O5. Lower valent vanadium oxides have the surface V=O species not only on the (010) plane but also the other crystalline planes, when the surface is slightly oxidized. In the present case, ESR and XPS spectra show that the addition of MOO, forms low-valent vanadium ions (Figures 3 and 4). Thus, the increase in the surface concentration of the redox sites can be attributed to the formation of V4+ ions due to the valence control in the V205-Mo03 solid solution. Mo/(V Mo) = 0.3 and 0.5. Broken lines in Figure 12 represent the calculated surface concentration of vanadium ions. Since the surface concentration of the redox sites was lower than the broken lines below the surface Mo content of ca. 0.4, the increase in the redox sites can be attributed to the formation of the surface V=O species on various planes, as mentioned above. However, above the surface Mo content of 0.4, the surface concentration of the redox sites exceeded the broken lines representing the concentration of surface vanadium ions. The result can be explained if the oxygen species bonded to the surface molybdenum ions are active for the reduction of N O with NH,. This was confirmed by the SIMS experiment as shown in Figure 6b. The spectrum indicates that the oxygen species bonded to molybdenum in V205-Mo03 catalyst were substituted with gaseous oxygen, or, in other words, the oxygen species also act as the redox sites during the reduction of N O with NH,. Probably, the surface Mo=O species act as the redox sites, because it is well-known that the surface Mo=O is active for the similar reaction steps, Le., the oxidative hydrogen abstraction and the reduction of oxygen, 19,26,27 MOO, itself is inactive in the NARP experiment, and a temperature higher by ca. 130 K than that on VZO5 is necessary for reduction of N O with NH, on Thus, the activation of the oxygen species bonded to the surface molybdenum ions may be due to the strong interaction with neighboring vanadium ions in the surface. The increase of the surface V=O species as well as the activation of the surface Mo=O species could be attributed to the mixing of the surface vanadium and molybdenum ions. The highest concentration of the surface redox sites was observed from Mo/(V + Mo) = 0.3 to 0.5. As discussed above, the structural feature of these catalysts is the atomic-scale mixing of vanadium and molybdenum ions in the bulk and on the surface of the catalysts. The trend of the surface concentration of the redox sites is similar to the degree of the mixing of the surface cations evaluated by SIMS (Figure 7) and the amount of the intermediate compounds evaluated by XRD (Figure 1). Therefore, the mixing of the surface vanadium and molybdenum ions should be the most important factor for the increase of the surface redox sites and the activation of the surface M e 0 species. A similar effect was reported for V205-WO, ~ata1ysts.l~ Thus, it was also concluded that the activation of the surface W=O species occurs by the effect of the mixing of surface vanadium and tungsten ions. However, a remarkable difference in the bulk structure was observed between V205-Mo0, and V205-W0, catalysts. Although no compounds were observed in the latter, a clear evidence for the intermediate compounds was obtained

+

(24) Anderson, A. J. Solid State Chem. 1982, 42, 263. (25) Miyamoto, A,; Hattori, A,; Murakami, Y . J . Solid State Chem. 1983, 47, 313.

(26) Hall, W. K.; Jacono, M. L. Proc. 6th Inr. Congr. Catal. 1977, I , 246. (27) Volta, J. C.; Desquesnes, W.; Moraweck, B.; Coudurier, G. React. Kinet. Catal. Lett. 1979, 12, 241. Tatibouet, J . M.; Germain, J. E.; Volta, J . C. J . Catal. 1983, 82, 240. Volta, J. C. Proc. Inr. Congr. Catal. 1984, 8, IV-451. Volta, J. C.; Tatigouet, J. M. J . Catal. 1985, 93, 467.

J . Phys. Chem. 1989, 93, 1490-1497

1490

in the former. It should be added that the M e 0 species are activated only on the intermediate compounds as shown below. Mol( V M o ) = 0.7 and 0.9. Since MOO, has redox activity, it is reasonable that the redox activity of the surface Mo=O species are enhanced by the effect of neighboring vanadium ions. However, the effect was less significant at high Mo content. The surface concentration of redox sites decreased above the surface Mo content of 0.6, where MOO, phase is the most abundant in the bulk. Since MOO, is inactive for the reduction of N O with NH, at this temperature (Figure 1l), the low concentration of redox sites at high Mo content may be due to the exposure of the inactive MOO, crystalline surface. Thus, well mixing of vanadium and molybdenum ions in the bulk, Le., the formation of the intermediate compounds, is required for the activation of the surface Mo=O species in the case of V205-Mo03 catalysts. Catalytic Activity and Active Sites. It has been reported that the activity of V 2 0 5 catalyst was increased by the addition of MOO, in the oxidation of benzene,2 butene: pr~pylene,~ and.CO3. The maximum activity in these reactions has been observed at 30 mol% MOO,, which is the limit of the formation of solid solution. Tarama et al.,,’ suggested that the high activity of V205-Md3 catalysts is due to the qualitative change in the active sites, Le., weakening of the V = O bond. However, the high activity of V20S-Mo0, catalysts has not been discussed in terms of the number of the surface V=O species, because of the lack of the method to measure the active sites. As shown in Figure 12, it was found in this study that the surface concentration of the redox sites is increased by the addition of MOO, to V205. The result agreed well with the above-mentioned increase in the activity of V205 catalyst by the addition of MOO,. It is quite reasonable that the catalytic activity depends on the surface concentration of active sites. Although the highest activity was frequently observed at 30 mol % MOO, (Mo/(V + Mo) = 0.181, the maximum concentration of the surface redox

+

+

sites was obtained at Mo/(V Mo) = 0.5. As shown in Figure 1, the solid solution of V2O5 is the abundant phase below 30 mol % (Mo/(V Mo) = 0.1 in this work), and intermediate compounds are abundant above 30 mol % (in this work, Mo/(V + M3) = 0.3 and 0.5). This may suggest that the activity of the surface redox sites on intermediate compounds is lower for the oxidation of hydrocarbons than that on solid solution. Acid sites do not seem to be responsible for the improvement of catalytic activity, because the type, strength, and surface concentration of acid sites are not varied by addition of MOO,. Thus, the above-mentioned increase in the activity of V205-MOO, catalysts may be attributed to not only the qualitative change of the surface V=O species but also the increase in the surface concentration of the V=O species on solid solution.

+

Conclusions The surface active sites of V2O5 catalyst under the promoting effect of MOO, were elucidated as follows. 1. The surface concentration of redox sites is increased by addition of MOO, to V205 catalysts. The surface V=O species are newly formed on the various crystal plane of solid solution, and the surface Mo=O species on the intermediate compounds are activated under the effect of V2O5 to act as the redox sites. 2. The acid sites of V2O5 catalyst are not modified significantly by the addition of MOO,. 3. The high activity of V205-Mo03 catalysts reported in the literatures may be attributed to the increase in the surface concentration of the V=O species on solid solution. Acknowledgment. This work was partially support by a grant-in-aid from the Ministry of Education, Science and Culture, Japan (No. 57470055). Registry No. MOO,, 1313-27-5; V,05, 1314-62-1;NO, 10102-43-9; NH,, 7664-41-7.

NMR Study of the Location of Bromide Ion and Methyl Naphthalene-2-sulfonate in Cationic Micelles: Relation to Reactivity Radu Bacaloglu,* Clifford A. Bunton,* Giorgio Cerichelli,’” and Francisco Ortegalb Department of Chemistry, University of California, Santa Barbara, California 931 06 (Received: April 29, 1988; In Final Form: August 24, 1988)

The ’H and 13Cchemical shifts of methyl naphthalene-2-sulfonateand naphthalene-2-sulfonateanion show that these compounds are located at the micellar interface in aqueous cetyltrialkylammonium bromides (alkyl = Me, Et, n-Pr, n-Bu). Increase of the surfactant headgroup size moves the ester closer to the cationic center. Reaction with Br- is faster and there is a more favorable interaction of the naphthalene r-system with the positive center. Micellar incorporation of Br- markedly increases the NMR spectral line width of 79Br,due to a disruption of hydration that is largest for the bulkiest surfactant headgroups. Bulkier headgroups decrease the halide ion concentration at the micellar surface, but this effect on overall reaction rate is offset by disruption of anion hydration.

We have examined effect of headgroup size in cetyltrialkylammonium chloride and bromide micelles (alkyl = Me, Et, n-Pr, n-Bu) on the reaction of methyl naphthalene-2-sulfonate (MeONs) with the corresponding halide anions:

a

~s:32~ 6

5

(1) (a) Present address: Centro C . N. R. di Studio sui Meccanismi di Reazione c/o Dipt. di Chimica Universita ‘La Sapienza” 00185 Rome, Italy. (b)-Present address: Department of Physical Chemistry, Facultad de Ciencias Quimicas, Universidad Complutense, Madrid, Spain.

0022-3654/89/2093-1490$01.50/0

1

4

MeONs

5

4

-0Ns

X = C1, B r

Although an increase in the size of the alkyl group decreases the 0 1989 American Chemical Society