Metal-TiO2 Catalysts - ACS Symposium Series (ACS Publications)

Feb 10, 1986 - DOI: 10.1021/bk-1986-0298.ch020. ACS Symposium Series , Vol. 298. ISBN13: 9780841209558eISBN: 9780841211308. Publication Date ...
0 downloads 0 Views 1MB Size
20 Metal-TiO2 Catalysts Electronic Effects During H2 Chemisorption, CO-H2 Interactions, and Photocatalysis

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on March 7, 2016 | http://pubs.acs.org Publication Date: February 10, 1986 | doi: 10.1021/bk-1986-0298.ch020

J.-M. Herrmann Equipe photocatalyse, C.N.R.S., Ecole Centrale de Lyon, B.P. 163, 69131 Ecully Cedex, France The electronic behaviour of Pt, Rh and Ni/TiO2 catalysts was followed by measuring in situ the conductivity σ of the support. From σ variations at room T, it is deduced that the metal is always enriched in free electrons from the support and that H2 chemisorption is followed by Η spill over on the oxide. On Pt/TiO2, σ measurements sho­ wed (i) that CO chemisorbs as a donor molecule, which can explain why an excess of electrons in the metal can counteract CO chemisorption and (ii) that CO does not dissociate on TiO2 or at the interface. At reaction tem­ perature (290°), CO does not dissociate either, which can hint that methanol is a primary product and, on adding H2, the decrease of σ shows that TiO2 surface is reoxidized by the oxygenated products (methanol, water), thus partly destroying SMSI. The electron exchange between the metal and its support play a fundamental role in SMSI as clearly demonstrated in the photocatalytic cyclopentane-deuterium isotopic exchange reaction. Recently t i t a n i a appeared as a non-conventional support for noble metal c a t a l y s t s , since i t was found to induce perturbations i n t h e i r H2 or CO adsorption capacities as well as i n t h e i r c a t a l y t i c a c t i v i ­ t i e s . This phenomenon, discovered by the EXXON group, was denoted "Strong Metal-Support Interactions' (SMSI effect) (1) and l a t e r ex­ tended to other reducible oxide supports (2), Two symposia were de­ voted to SMSI, one i n Lyon-Ecully (1982) (3) and the present one i n Miami (1985) (j4) and presently, two main explanations are generally proposed to account for SMSI: ( i ) either the occurence of an e l e c t r o ­ nic effect (2 5-13) or ( i i ) the migration of suboxide species on the metal p a r t i c l e s (14-20), The second hypothesis was e s s e n t i a l l y i l l u s t r a t e d on model catalysts with spectroscopic techniques,It can be noted that both p o s s i b i l i t i e s do not necessarily exclude each other and can be considered simultaneously (21). In the present paper, the metal-support electronic interactions i n various metal catalysts-mainly Pt - were followed by measuring i n s i t u the e l e c t r i c a l conductivity of the s o l i d s either i n the "nor­ mal" or the "SMSI" state, when i n contact with various atmospheres (vacuum, H2, 02> (CO+H2). The manifestation of the electronic factor 1

y

0097-6156/ 86/0298-0200506.00/0 © 1986 American Chemical Society

Baker et al.; Strong Metal-Support Interactions ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

20.

HERRMANN

Metal- Ti0 Catalysts: Electronic Effects

201

2

during c a t a l y s i s was i l l u s t r a t e d by the photocatalytic change between deuterium and cyclopentane.

isotopic

ex-

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on March 7, 2016 | http://pubs.acs.org Publication Date: February 10, 1986 | doi: 10.1021/bk-1986-0298.ch020

Experimental Catalysts;The three types of catalysts (Pt, Rh, Ni/Ti02) where prepared by impregnating a desired quantity of titanium dioxide (anatase Degussa P-25; 50 m2 g-1) with aqueous solutions of compounds containing the cations of the metal chosen (chloroplatinic a c i d , rhodium t r i c h l o r i d e and n i c k e l hexammine n i t r a t e ) at concentrations appropriate to y i e l d the proper metal loading (5 wt% unless otherwise stated). The impregnated s l u r r i e s were subsequently evacuated at 80 C i n a rotating f l a s k , dried at 110°C for 2 h, reduced at 480 C overnight i n hydrogen, cooled down to room temperature under nitrogen flow and kept i n v i a l s u n t i l l further use. Chemical analyses gave metal loadings of - 5 wt%, The catalysts were characterized by hydrogen chemisorption and transmission electron microscopy (TEM), Pt and Rh were present as homodispersed p a r t i c l e s of 2 and 3,5 nm respectively, homogeneously distributed on a l l the quasi-spherical (d^25 nm) p a r t i c l e s of P-25 Degussa anatase. Concerning Pt, d i f f e r e n t catalysts were prepared with loadings varying from 0,1 to 10%: the p a r t i c l e size remained unchanged (1,5-2 nm) even with a metal loading varying by 2 orders of magnitude (22), Nickel p a r t i c l e s were far bigger and because of a lack of contrast could not be seen by TEM, Magnetic measurements, however, yielded a mean size of about 13,5 nm, A rough estimation showed that only 10 to 20% of T102 p a r t i c l e s were i n contact with a n i c k e l one. e

e

E l e c t r i c a l conductivity measurements: They were carried out i n a s t a t i c - t y p e c e l l , designed for powder samples, which allows i n s i t u measurements from the beginning of the pretreatment up to subsequent solid-gas interactions. The r e s u l t s refer only to the support i n the case of supported metal c a t a l y s t s , The photocatalytic isotopic exchange between cyclopentane and deuterium (CDIE) was carried out i n a s t a t i c fused s i l i c a photoreactor described i n ref (23), Results and Discussion Effect of SMSI on H? chemisorption: The various M/T102 catalysts were reduced i n 250 Torr H2 either at low temperature (200 C; LTR samples) or at high temperature (500 C; HTR samples), LTR Ni/Ti02 sample was exceptionnally reduced at 300 C to make sure that a l l n i c k e l was completely reduced, H2 adsorption capacities are l i s t e d in Table I. From Table I, i t can be seen that for a constant metal loading (5%), the s e n s i t i v i t y to SMSI of the present samples varies as Ni » Pt>Rh, This c l a s s i f i c a t i o n i s purely q u a l i t a t i v e and not i n d i cative of the proper influence of the nature the metal since several parameters such as texture and dispersion can influence the extent of SMSI, For instance, n i c k e l i s under the shape of far larger part i c l e s than Pt or Rh ones with 2 to 3 times more atoms. On the cont r a r y , in the case of Pt c a t a l y s t s , the comparison between catalysts i s more meaningful since the metal c r y s t a l l i t e size i s homodispersed e

e

e

Baker et al.; Strong Metal-Support Interactions ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

STRONG METAL-SUPPORT INTERACTIONS

202

and constant. It can be observed from Table I ( i ) that f o r two d i f ­ ferent preparations with the same Pt content (5%), the reproducibi­ l i t y i s correct and ( i i ) that f o r two d i f f e r e n t loadings (0,5 and 5 P t % ) , the effect of SMSI on the adsorption i s stronger for the smaller one and i n i t i a t e s at lower temperatures (300 C), This beha­ viour w i l l be observed and discussed further i n the c a t a l y t i c section. e

Table I. Adsorption of H

on various M/TiO

?

catalysts

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on March 7, 2016 | http://pubs.acs.org Publication Date: February 10, 1986 | doi: 10.1021/bk-1986-0298.ch020

Amount of chemisorbed

(ymol/g,cat)

Reduction temperature 200 Pt/TiO Pt/TiO^ Pt/TiO^ Rh/Tio' Ni/TiO^

5% 5% 0,5% 5% 5%

300

41

500

41 3

8

6,5 67 -

6 4 0 24 0

-

3 8

E l e c t r i c a l conductivity study of M/TiO^ catalysts This study has been previously described i n ref (11,24)and can be summarized as follows, e

(i) LTR samples (Tg « 200°C): In H2 at 200 C, since the presence of a metal deposited increases the conductivity of t i t a n i a by 1 to 2 orders of magnitude, i t was inferred that the metal catalyzes the reduction of the oxide with s p i l t over atomic hydrogen which f i r s t creates OHg^surface groups: H

2

+

M

2

s "

S "

M

H

+

2

H

V M

°s~

s

( 1 )

+

0H

s"

+

e

"

( 2 )

whose dehydration forms anionic vacancies VQ22 0H~ &

2

H 0(g) + 0 " + V 22

Q

(3)

which are generally singly ionized at the temperatures considered V 2Q

5=?

V 2- + e"

(*)

Q

Simultaneously, because of the good physical and thence e l e c t r i ­ cal contact between both s o l i d phases, there i s an alignment of their Fermi l e v e l s which demands an electron migration - even limited from the support to the metal i n agreement with the respective va­ lues of the work functions of the metal and of the reduced support, e~ + M ^ e"~ (5) The electron transfer of Eq (5) i s confirmed when H2 i s evacua­ ted at 400 C since the e l e c t r i c a l conductivity of M/Ti0 , / «t decreases whereas increases because of the creation or new anionic vacancies due 2to the increase of temperature, e

a

2

2

Ο " φ*

1/2 0 ( g ) 2

+ V £ _ + e" Q

M

T i 0

(6)

In Eq(6), the charge balance i s obtained by summation i n both sides of the equation of a l l the charges mentioned, included those

Baker et al.; Strong Metal-Support Interactions ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

20.

HERRMANN

203

Metal- Ti0 Catalysts: Electronic Effects 2

in subscripts ( l i k e VQ2-} i n order to account for the fact that spe­ cies such as 0^~ and V Q 2 - are neutral with respect to the s o l i d , Eq(5)is also confirmed when Τ i s decreased under vacuum from 400°C to room temperature since M/T102 samples remain semiconductors with a constant activation energy of conduction E i n contrast with bare t i t a n i a which behaves as a quasi-metallic conductor ( E - 0), F i n a l l y , the introduction of hydrogen at room temperature on M/T102 produces a reversible increase of QM/Ti02 which follows the isotherm law: . σ = a + b P__ (7) 2 which can be accounted f o r by Eqs (1), (2) and (5), ( i i ) HTR samples (T = 500 C), SMSI state: Q u a l i t a t i v e l y , M/Ti0 samples behave as tne LTR ones. However, as evidenced by the higher value of σ, the reduction of t i t a n i a i s much more important, A T l ^ y phase was even i d e n t i f i e d (18,25), Consequently, the r e l a t i v e e l e c ­ tron enrichment of the metal i s stronger i n agreement with the alignment of the Fermi l e v e l s of the metal and of reduced t i t a n i a whose work function decreases with the reduction l e v e l . This electron excess i n the metal, even i f i t i s l i m i t e d , i s thought to be at the o r i g i n of the SMSI effect which p a r t l y suppresses H chemisorption, especially i f t h i s chemisorption occurs with the creation of dipoles at the surface of metal (26,27,28), The absence of chemisorptive pro­ perties under SMSI conditions f o r catalysts having a low loading, H h i c h depends on the nature and texture of the metal (0,5% for Pt and 5% for N i , Table I) i s confirmed by the absence of hydrogen s p i l l over (do/dP = 0 i n Eq,(7))on these samples, whereas f o r a l l other samples whose H2 chemisorption capacity i s not n i l when i n the SMSI state, the relationship of Eq 7 i s duly observed. The restauration of a normal state by exposure to oxygen i s ex­ plained by the reoxydation of the support with the f i l l i n g of sur­ face anionic vacancies (Eqr6) which increases t i t a n i a ' s work func­ tion and requires the retrocession of excess electrons to the oxide, c >

c

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on March 7, 2016 | http://pubs.acs.org Publication Date: February 10, 1986 | doi: 10.1021/bk-1986-0298.ch020

1

/

2

e

2

2

CO + H interactions on 5 wt% P t / T i Q 2

2

i n the SMSI state

Interactions at room temperature: When CO i s f i r s t introduced ( F i g . l ) , σ increases instantaneously and then remains independent of P . The fact that σ does not decrease means that CO does not d i s ­ sociate on t i t a n i a nor at the interface, otherwise the f i l l i n g of anionic vacancies by atomic oxygen (Eq,-6) would have decreased substantially σ by consuming free electrons. The sharp i n i t i a l i n ­ crease, on the contrary, shows that CO chemisorb on Pt with a donor effect probably due to the creation of dipoles as proposed f o r chemisorption which renders ohmic the e l e c t r i c a l contact between the metal and i t s semiconductor support (26, 17, 28)According to these authors, the creation of a dipole layer decreases the work function 0 of the metal which approaches the electron a f f i n i t y of the semi­ conductor, thus suppressing the Schottky b a r r i e r . Presently CO ad­ sorbs as a donor molecule on Pt decreasing 0 , which allows e l e c ­ trons to be restituted to t i t a n i a . The absence of variations of σ versus Ρ f o r subsequent increasing CO pressure means that the sur­ face i s already saturated at these pressures. The introduction of H i n the presence of CO increases σ but more slowly than did the i n i t i a l dose of CO, The k i n e t i c s i s compa­ rable to that of hydrogen s p i l l o v e r which can be observed on t h i s r

M

?

Baker et al.; Strong Metal-Support Interactions ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

STRONG METAL-SUPPORT INTERACTIONS

204

catalyst since chemisorption under SMSI conditions i s not n i l (Table I ) , This means that i s able to displace CO from certain of i t s s i t e s (29) from which i t can subsequently migrate on the sur­ face of t i t a n i a . This requires that these adsorption s i t e s are the weakest for CO and close to the interface to l e t hydrogen s p i l l over. Interactions at reaction temperature ( 2 9 0 ° C ) : When the same ex­ periments are repeated at 290 C ( F i g , 2 ) , CO does not dissociate on t i t a n i a since σ does not decrease but on the contrary conserves i t s donor character to platinum which retrodonates some excess electrons to the support. The introduction of H increases f i r s t sharply σ for a few se­ conds and then produces a rather slow decrease. This behaviour shows that Η i s always-able to adsorb d i s s o c i a t i v e l y on a surface satura­ ted with CO but now the s p i l l over i s prevented - or at least mas­ ked - by the reaction i t s e l f whose oxygen - containing products, mainly methanol and water, are able to p a r t i a l l y reoxidize the sur­ face of t i t a n i a by f i l l i n g some of i t s anionic vacancies, A similar decrease of σ can be obtained by introducing pure water. The sharp i n i t i a l increase of σ observed i n Fig,2 i s to be ascribed to a ther­ mal effect due to the starting reaction and the subsequent exother­ mic reoxidation of t i t a n i a . The drop of σ induced by the p a r t i a l reoxidation of t i t a n i a s surface i s small i n comparison with the reoxidation by gaseous oxy­ gen at room temperature (11,24).This i s not unexpected since (i) oxygen i s highly e l e c t r o p h i l i c and r e a d i l y chemisorbs on reduced t i t a n i a ; ( i i ) the temperature of reoxidation by 0^ i s by 290°C lower than for (CO + H ) , which does not unfavor the strongly exothermic reoxydation of T i 0 but substantially decreases σ because TiO« has recovered i t s complete semiconductor behaviour with a high a c t i v a ­ t i o n energy of conduction and ( i i i ) i n the case of (CO + H ) , the atmosphere remains highly reducing, permitting the maintenance of anionic vacancies with free electrons of conduction. In f a c t , the σ l e v e l at the end of (CO + Η ) interactions at 290°C i s comparable to that of P t / T i 0 under H at 200°C i n the "normal s t a t e " . Consequently, although CO and H chemisorptions are decreased by a factor of 5 and 9 respectively on the 5% Pt sample used, the absence of i n h i b i t i o n for CO + HL reaction can be accounted for by two simultaneous explanations: (i) the active CO and H species are only those whose chemisorption sites are unaffected by SMSI (see r e f , (30) and close to the interface perimeter (31), and/or ( i i ) SMSI does not exist any longer during the (CO + ÏÏ7) reaction, since the catalyst returns to the "normal" state by p a r t i a l reoxidation of the support, In t h i s last case, the exothermicity of t i t a n i a * s reoxidation can increase the catalyst temperature and accelerate the i n i t i a l reaction rate,

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on March 7, 2016 | http://pubs.acs.org Publication Date: February 10, 1986 | doi: 10.1021/bk-1986-0298.ch020

e

2

f

2

2

2

2

2

2

Photocatalytic cyclopentane-deuterium isotopic exchange (CDIE) This reaction was chosen to i l l u s t r a t e the occurence of an e l e c tronic factor i n a c a t a l y t i c test under SMSI conditions. This react i o n appears to be p a r t i c u l a r l y suitable for t h i s purpose since(i) the photonic a c t i v a t i o n involves the formation of photoelectrons and photoholes and ( i i ) the absence of oxygen-containing molecules as well i n the reactants as i n the products allows to avoid the destruction of the SMSI state by a p a r t i a l reoxidation of the support.

Baker et al.; Strong Metal-Support Interactions ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on March 7, 2016 | http://pubs.acs.org Publication Date: February 10, 1986 | doi: 10.1021/bk-1986-0298.ch020

HERRMANN

Metal-Ti02 Catalysts: Electronic Effects

T = 563 κ

2 h

1 0 0

t/ min

Figure 2 . (CO + H ) interactions on P t / T i 0 i n the SMSI state at reaction temperature (290°C) (Pressures i n T o r r ) . 2

2

Baker et al.; Strong Metal-Support Interactions ACS Symposium Series; American Chemical Society: Washington, DC, 1986.

206

S T R O N G METAL-SUPPORT

T H E REACTION MECHANISM HAS BEEN DESCRIBED TION I S CARRIED ACTIVITY LIGHT SE

OUT AT TEMPERATURES

(PT,NI)

I S REQUIRED OTHERWISE,

CLINES

AND STOPS

TO CONFER

OF NON-RENEWABLE

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on March 7, 2016 | http://pubs.acs.org Publication Date: February 10, 1986 | doi: 10.1021/bk-1986-0298.ch020

GROUPS

FACT

NAKED T I T A N I A

OF

I S TOTALLY POINT THAT

THAT T H E HIGHER

LEVELS

TANCE OF T H E SAMPLE

I T HAS BEEN SHOWN

UNDER VACUUM T H E ELECTRONS

CORRESPONDING

TO T H E ILLUMINATED

OF THE METAL I S SUPPORTED

(33)

TRAPPED -

ISOTHERM

(Σ « A + B ^ ^

I N T H E DARK ( E Q , 7 )

ALL

UNDER H ^ , THERE

STILL

,

THESE OBSERVATIONS

)

FORMALLY " A " BEING

IDENTICAL

TO THAT O B ­

PROPORTIONAL

ENABLED US TO PROPOSE

SELECTIVE

DISSOCIATIVE D, 2

TO T H E

CHEMISORPTION

+ 2 PT ^ S

S

H

CYCLIC M E ­

OF D « ON PLATINUM

2 PT - D S

~ S

10

A 8-STEP

FORMATION OF MONODEUROCYCLOPENTANE,

AND R E V E R S I B L E WEAK CHEMISORPTION

2°/

2

BUT WITH

EXISTS A

BY A PHOTOCON­

FLUX.

CHANISM FOR T H E 100% 1°/

BY T H E

T H E HIGHER T H E

AND T H E SMALLER T H E PHOTOCONDUC­

AND ( I I )

R E V E R S I B L E S P I L L OVER OF ATOMIC HYDROGEN, DESCRIBED

RADIANT

DEHYDROXYLATED

PHOTO-INACTIVE,

T H E NUMBER OF PLATINUM P A R T I C L E ,

NUMBER OF PHOTO-ELECTRONS

TAINED

I D E N T I F I E D AS

EITHER

OF V I E W ,

(I)

SUPPORT D E ­

I N T H E SUPPORT ARE ATTRACTED BY THE METAL BECAUSE OF

T H E ELECTRON ENRICHMENT

DUCTIVITY

PHA­

OF A METAL

TO AN EXHAUSTION

T H E S E S I T E S HAVE BEEN

FROM T H E ELECTRONIC

THE ALIGNMENT OF T H E FERMI STATE -

OUT ON T H E BARE

SINCE

BY PHOTOCONDUCTANCE MEASUREMENTS PHOTO-PRODUCED

T H E PRESENCE

A C E R T A I N TIME CORRESPONDING

I N H ^ INSTEAD

MOREOVER,

HOWEVER,

A C A T A L Y T I C CHARACTER TO T H E REACTION

ACTIVE S I T E S .

DEUTERATED HYDROXYL OR PRETREATED

I T ONLY OCCURS WHEN N E A R - U V

T H E REACTION CARRIED AFTER

THE REAC­

THUS SHOWING THAT T H E A C T I V E

I S CONSTITUTED BY T H E SUPPORT.

SINCE,

IN R E F , ( 2 3 ) ,

< - 1 0 ° C WHERE NO DARK THERMAL

OF T H E METAL CAN B E D E T E C T E D .

I S ADMITTED ONTO T H E S O L I D ,

INTERACTIONS

«10

OF CYCLOPENTANE ON T I T A N I A ,

(ads)

CREATION OF ELECTRON-HOLE PAIRS BY U V - L I G H T ( T I 0 ) + HV - E - + P+ ( H V ^ Ε = 3 EV) G MIGRATION OF PHOTOELECTRONS TO PLATINUM O

2

3°/

E" 4°/

+ PT ^

E ^

REACTION OF PHOTOHOLES WITH N E G A T I V E L Y 0D~

+ P

-

+

* 5 ° / D E A C T I V A T I O N OF OD PENTANE MOLECULE 8

0 D

6°/ PORT Ί IKE

CHARGED OD

GROUPS

OD * S

S

+

REVERSE

C

5

H

I0


^