Chemical vapor deposition of trimethylaluminum on dealuminated faujasite zeolite

Research Article

pubs.acs.org/acscatalysis

Chemical Vapor Deposition of Trimethylaluminum on Dealuminated

Faujasite Zeolite

Evgeny A. Pidko,* Sami M. T. Almutairi, Brahim Mezari, Pieter C. M. M. Magusin,

and Emiel J. M. Hensen*

Schuit Institute of Catalysis, Laboratory of Inorganic Materials Chemistry, Eindhoven University of Technology, Den Dolech 2, 5600

MB Eindhoven, The Netherlands

*

S Supporting Information

ABSTRACT: Chemical vapor deposition of trimethylalumi-

num (TMA) was explored as an approach for the preparation

of model faujasite-type catalysts containing extraframework

aluminum. The decomposition of the grafted organoaluminum

species was investigated in hydrogen and oxygen atmosphere.

The process of grafting Al-containing species and the

associated changes of the zeolite hydroxyl groups were

followed by in situ FTIR spectroscopy. The state of intrazeolite

Al atoms, the changes in zeolite structure and acidity caused by

the CVD procedure as well as by subsequent treatment were

1

29

27

analyzed in detail by H, Si, and Al MAS NMR, CO_a_d_sIR,

H/D exchange of acidic hydroxyl groups with perdeuterobenzene, and propane cracking. Reaction of an extraframework

aluminum-free high-silica faujasite zeolite with TMA leads to nearly complete substitution of the bridging hydroxyl groups with

Al species. The reaction, however, does not produce uniform homogeneously distributed species. Because of the high reactivity of

TMA, the zeolite lattice is partially decomposed resulting in its partial dealumination and formation of stable Si-CH moieties.

3

The exact conditions of post-CVD treatment inﬂuence strongly the chemical and catalytic properties of the zeolites. The

strongest increase of the propane conversion rate was observed when grafted TMA species were decomposed in H at high

2

temperature. Such zeolite displays much higher activity per Brønsted acid site in propane cracking than a commercial

ultrastabilized Y zeolite. It is proposed that the activity enhancement is related to strong polarization of a fraction of the zeolite

Brønsted acid sites by Lewis acid sites formed by the hydrogenolysis of grafted TMA complexes.

KEYWORDS: faujasite, dealumination, chemical vapor deposition, trimethylaluminum, extraframework aluminum, propane cracking,

H/D exchange, NMR, acidity, infrared spectroscopy

1

. INTRODUCTION

mesoporosity. Even in simple test reactions, like cracking or

hydroconversion of alkanes, materials of similar composition

show hugely diﬀerent activities. The usual interpretation of

such diﬀerences involves the assumption that zeolitic acid sites

can diﬀer in strength. Nonframework aluminum species due to

partial framework dealumination can boost the activity of

Zeolites are extensively used in the oil reﬁning and

petrochemical industry as acid catalysts to crack carbon−

carbon bonds, isomerize and oligomerize alkenes and alkylate

aromatics. Zeolite acidity is understood to a signiﬁcant

degree.

tetrahedral aluminum species substituting silicon atoms in the

framework. Zeolite Y, the preferred zeolite in hydrocracking

catalysts, has the faujasite structure which features a single

crystallographic tetrahedral site. Ideally, at high Si/Al ratios

dilute Brønsted acid sites (BAS) in a pure faujasite material will

therefore be equivalent. After hydrothermal synthesis of zeolite

Y, its Al content is high (Si/Al = 2.5), while its stability and

1

,2

3

−5

Acid catalytic activity resides at sites involving

8

,9

zeolites in cracking reactions or aﬀect its selectivity. These

strongly Lewis acidic aluminum species do not seem to

contribute to protolytic cracking, but instead are believed to

10,11

enhance the strength of Brønsted acid sites in their vicinity

and to promote alternative alkane activation paths such as

12

6

dehydrogenation.

The extracted Al atoms form the commonly named

extraframework Al (EFAl) phase, which includes charged

complexes at cationic positions and neutral aggregates occluded

acidity are low. The low acidity is understood in terms of the

next-nearest neighbor model. According to this model,

7

13−15

in the micropore space.

Part of the EFAl phase may also

bridging hydroxyl groups in Al-rich regions of zeolites have

the strongest O−H bonds because of higher basicity of lattice

oxygens. In practice, a hydrocracking catalyst will contain

steam-calcined stabilized faujasites with Si/Al molar ratios in

the 4.5−50 range. The steam calcination procedure renders the

zeolite more stable and acidic and also creates additional

be located in the mesopores or at the external surface. Because

Received: March 7, 2013

Revised: April 25, 2013

©

XXXX American Chemical Society

1504

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

of the heterogeneous nature of the EFAl phase, it has not been

possible yet to unambiguously determine the structure of such

species nor their role in catalytic cracking reactions. The

formation of EFAl species in the zeolite pores is believed to

increase acidity. The nature of these EFAl species and how they

aﬀect Brønsted acidity of zeolites is an issue of continuous

debate. It has been postulated that these EFAl polarize

neighboring BAS and increase their activity in acid-catalyzed

precursor to the industrially important MAO cocatalyst for

41−55

oleﬁn polymerization.

The modiﬁcation of dealuminated Y

zeolite with TMA followed by impregnation of platinum has

been shown to result in enhanced reactivity of the zeolite

56−58

catalyst for decane dehydrogenation.

As a starting material, we employ a high-silica faujasite zeolite

prepared by selective substitution of Al with Si atoms via

F

treatment with ammonium hexaﬂuorosilicate (NH ) SiF of

4

2

6

1

6

7,35

reactions. In recent work, Iglesia and co-workers opposed this

view by suggesting that the activity enhancement is mainly

associated with the change of eﬀective void of the supercage

due to the presence of EFAl rather than to changes in intrinsic

zeolite Y.

EFAl species were then introduced by chemical

vapor deposition of TMA followed by the decomposition of the

grafted organoaluminum species under oxidizing or reducing

conditions. The state of EFAl in the prepared materials was

1

7

27

strength of BAS. On the contrary, van Bokhoven and co-

workers demonstrated that EFAl species in steam-activated

zeolite Y predominantly occupy cationic positions in the

sodalite cages and, therefore, cannot directly inﬂuence the

investigated by Al MAS NMR spectroscopy and FTIR

spectroscopy of adsorbed CO. Brønsted acidity of the zeolite

catalysts was investigated by monitoring the selective H/D

exchange of acidic hydroxyl groups with C

D

by IR

6

18

59−62

adsorption of hydrocarbons in the supercages.

The introduction of extraframework Al in low-silica X and Y

spectroscopy.

The acid catalytic activity of these zeolites

was determined by propane cracking experiments.

19

zeolite results in a substantial increase of acidity of BAS.

Solid-state NMR spectroscopy is a powerful method for the

investigation of the oxygen coordination, local symmetry, and

concentration of Al atoms at framework and extraframework

positions in zeolites. Many eﬀorts have focused on under-

standing the role of EFAl on BAS using diﬀerent NMR

2. EXPERIMENTAL SECTION

2

.1. Catalyst Preparation. Dealuminated AHFSY zeolite

was prepared according to the procedure described in ref 35

(

AHFSY-60) by reacting the ammonium form of zeolite Y

Akzo Nobel, Si/Al = 2.5) zeolite with an aqueous solution of

1

13

spectroscopic techniques such as H DQ-MAS NMR and

C

ammonium hexaﬂuorosilicate ((NH ) SiF , AHFS). The

4

2

6

1

3

20−22

MAS NMR of adsorbed 2- C-acetone.

It was also used to

framework Si/Al ratio in the resulting AHFSY material was

.2. Subsequent modiﬁcation with trimethylaluminum was

performed by chemical vapor deposition (CVD) of Al(CH₃)₃

TMA, Aldrich Chemicals, 95%) on dehydrated AHFSY

reveal the importance of the spatial proximity of Brønsted and

4

Lewis acid sites in dealuminated H−Y zeolite to enhance

2

3−25

26

27

acidity.

More recently, also Al DQ-MAS NMR

(

spectroscopy was employed to study the location of various

types of Al species in dealuminated Y zeolite. Besides, several

studies employed monomolecular cracking of hydrocarbons to

characterize the intrinsic reactivity of the zeolitic Brønsted acid

zeolite. Prior to the TMA CVD treatment, the AHFSY powder

was pressed, crushed, and sieved to a fraction between 250 and

500 μm. The catalyst was calcined in a mixture of 20 vol %

oxygen in nitrogen at a ﬂow rate of 100 N mL/min while

heating at a heating rate of 2 °C/min followed by an isothermal

period of 6 h to obtain the hydrogen form of AHFSY. The

dehydrated catalyst was transferred into a nitrogen-ﬂushed

glovebox without exposure to air. All subsequent manipulations

were performed under inert atmosphere. To a glass vessel

containing a small sample bottle with 1 g of dehydrated

AHFSY, 1 mL of TMA was added (see Supporting

Information). After CVD for 48 h, the material was evacuated

for 2 h to remove unreacted TMA and reaction products. The

modiﬁed zeolite was then kept under an inert atmosphere.

Before the catalytic tests, the TMA-modiﬁed samples were

27−34

sites of diﬀerent zeolites.

One of the reasons why it

remains diﬃcult to draw clear conclusions on the structure and

role of EFAl is the signiﬁcant structural heterogeneity of EFAl-

containing faujasites.

To understand better the nature and catalytic role of EFAl in

faujasite zeolites, it would be useful to develop model zeolites

containing well-deﬁned EFAl species. In our previous work, the

inﬂuence of extraframework Al in a model high-silica faujasite

(

Si/Al ≈ 5, no EFAl) introduced by conventional methods such

as incipient wetness impregnation and ion exchange was

3

5

investigated. Strong synergy in propane conversion was

observed between BAS and cationic EFAl species. Despite

eﬀorts to control the type of Al species, it was found that there

remains a considerable degree of heterogeneity in such model

samples.

either treated in pure H for 2 h at 550 °C and/or calcined in a

2

mixture of 20% O in He. The activated catalysts were denoted

2

as AHFSY-TMA-Red, AHFSY-TMA-Red-Oxd, and AHFSY-

TMA-Oxd, respectively, for the catalyst activated by reduction

in H , a cycle of reduction−oxidation and oxidation in O at

The surface organometallic approach involves modiﬁcation

of solid surfaces by catalytically active species by selective

2

550 °C.

36

reaction of its functional groups. In zeolites the hydroxyl

groups will react with volatile metal−alkyl species. This

approach has been explored before for preparation of model

2.2. Catalyst Characterization. X-ray Diﬀraction (XRD).

XRD patterns of zeolites were recorded on a Bruker D4

Endeavor Diﬀractometer using Cu Kα radiation with a

wavelength of 1.54056 Å. 2θ angles from 5° to 60° were

measured with a step size of 0.077° and a time per step of 1 s.

The catalysts were ground and pressed in sample holders for

measurements under ambient conditions. Crystallinity of the

zeolite samples was calculated from the XRD patterns using the

Topas software.

3

7,38

systems of Ga/ZSM-5.

It has also been used to modify

3

9

ZSM-5 with Zn cations. The reaction of Brønsted acid sites

and the highly reactive organometallic precursor results in well-

deﬁned cationic surface species, which can be further modiﬁed

by subsequent thermochemical activation. Herein, the use of

volatile trimethylaluminum (Al(CH ) , TMA) is explored as a

3

potential precursor for well-deﬁned EFAl species in zeolite Y.

Previously, it has been demonstrated that TMA can be

eﬀectively incorporated into mesoporous silica MCM-41 via

Elemental Analysis. The aluminum content in zeolite

catalysts was determined by elemental analysis, which was

carried out by ICP-OES (Spectro Ciros CCD ICP optical

emission spectrometer with axial plasma viewing). For the ICP

40

various postsynthesis procedures. TMA itself is a prototypical

1

505

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

measurements the samples were dissolved in a 1.5 mL solution

of HF/HNO /H O (1:1:1) acid mixture.

temperature was reached an initial spectrum was recorded. CO

was dosed via a sample loop connected to a six-way valve (50

μL). FTIR spectra were recorded by accumulating 64 scans at a

resolution of 2 cm . Diﬀerence spectra were obtained by

subtracting the initial spectrum of the dehydrated catalyst from

the spectra obtained at increasing CO coverage.

3

2

Argon Porosimetry. Surface area and porosity of the zeolites

were determined by argon physisorption. In a typical

experiment, a dehydrated zeolite sample (∼100 mg) was

−1

loaded in the sample holder inside a N -ﬂushed glovebox and

2

outgassed at room temperature prior to the sorption measure-

ment. The measurements were performed in static mode at

More detailed descriptions of the characterization procedures

can be found in refs 34, 39.

−

186 °C on a Micromeritics ASAP-2020 apparatus. The

Brunauer-Emmett−Teller (BET) equation was used to

Solid-State NMR. NMR experiments were performed in a

magnetic ﬁeld of 11.7 T on a Bruker Avance DMX500

1

29

calculate the speciﬁc surface area from the adsorption data

operating at 500 MHz for H, 99 MHz for Si, and 132 MHz

for ²⁷Al. The NMR measurements were carried out using a 4-

mm MAS probehead with a sample rotation rate of 12.5 kHz

obtained in the p/p range of 0.05−0.25. The volume of

0

mesopores was calculated using the Barrett−Joyner−Halenda

1

27

29

(

BJH) method on the adsorption branch of the isotherm.

Fourier Transform Infrared Spectroscopy (FTIR). FTIR

for H and Al, and 10 kHz for Si NMR measurements.

1

H NMR spectra were recorded with a Hahn-echo pulse

spectra were recorded with a Bruker Vertex V70 FTIR

spectrometer in transmission mode. Typically, about 10 mg

sequence p1-τ1-p2-τ2-aq with a 90° pulse p1 = 5 μs and a 180°

p2 = 10 μs. The interscan delay of 120 s was chosen for

of zeolite sample was pressed into a self-supporting wafer

29

quantitative spectra. Quantitative Si NMR spectra were

2

(

density of 5.0−6.0 mg/cm ) and placed in a controlled-

recorded using a High Power proton Decoupling direct

environment transmission cell. The spectra were normalized to

the thickness of the wafer, determined from the weights of the

sample in the dehydrated state. Prior to recording spectra, the

excitation (DE) pulse sequence with a 90° pulse duration of

1

29

5

(

μs and an interscan delay of 360 s. H- Si cross-polarization

CP) spectra were obtained using an interscan delay of 3 s and

zeolites were dehydrated in an O ﬂow while heating from

27

2

a contact time of 5 ms. Al NMR spectra were recorded with a

single pulse sequence with a 18° pulse duration of 1 μs and a

interscan delay of 1 s. MQMAS spectra were recorded by use of

the three-pulse sequence p -t -p -τ-p -t for triple-quantum

room temperature to 550 °C at a heating rate of 5 °C/min.

After an isothermal period of 1 h at 550 °C, the sample was

evacuated for 50 min at 550 °C and cooled to room

temperature in vacuum.

1

2

3 2

generation and zero-quantum ﬁltering (strong pulses p1 = 3.4

μs and p2 = 1.4 μs at v1 = 100 kHz; soft pulse p3 = 11 μs at v1

The reaction of TMA with zeolite Brønsted acid sites and the

subsequent conversion of the grafted species was monitored in

situ by FTIR spectroscopy. The zeolite AHFSY wafer was

loaded in the cell, dehydrated and evacuated followed by a

short exposure to TMA vapor at room temperature. After a

hold time of 20 min, the sample was evacuated for 2 h and

subsequently either reduced (H ) or oxidized (O ) at a

1

27

=

8 kHz; ﬁlter time τ = 20 μs; interscan delay 0.2 s). H-{ Al}

TRAnsfer of Population in DOuble Resonance (TRAPDOR)

spectra were recorded by use of a 90°-τ -180°-τ proton pulse

1

2

sequence with equal time intervals τ = τ = 795 μs and with

1

2

27

irradiation at the Al NMR frequency, during τ . The same

experiment without Al irradiation served as a blank

1

2

27

pressure of 50 mbar and increasing temperatures (100, 200,

00, 400, and 500 °C) for 30 min followed by evacuation at

experiment for reference. The interscan delay was 10 s.

3

1

Tetramethylsilane (TMS) was used for calibrating the H and

each temperature setting and recording an FTIR spectrum.

Acidity Characterization by H/D Exchange with Perdeu-

terated Benzene. The intrinsic acidity of diﬀerent faujasite

zeolites was probed using a methodology based on the selective

H/D exchange of hydroxyl groups in the zeolite upon the

29

27

Si NMR shift, and a concentrated Al(NO ) solution for Al

3

NMR shift calibration.

The diﬀerent TMA modiﬁed AHFSY zeolites were loaded in

a 4-mm zirconia NMR rotor in a nitrogen-ﬂushed glovebox and

tightly closed with a Kel-F cap. After that, the rotor was

transferred to the NMR probe under nitrogen gas atmosphere.

59−62

reaction with perdeuterated benzene.

The progress of the

H/D exchange reaction was monitored in situ by infrared

spectroscopy. FTIR spectra were recorded with a Bruker Vertex

V70v FTIR spectrometer in transmission mode following the

procedure described in ref 35. Perdeuterobenzene (C D ,

2

.3. Catalytic Activity Measurements. Similar to our

35

previous study, the acid activity of the zeolites was

determined by measuring the rate of monomolecular cracking

of propane. Catalytic activity measurements were performed in

an atmospheric-pressure single-pass quartz microﬂow reactor at

6

Aldrich, purity 99.96%) was dosed into the cell containing a

dehydrated zeolite sample from a glass ampule via a computer

5

90 °C. The reactor was loaded with the catalyst under inert

atmosphere. The feed mixture was delivered by thermal mass

ﬂow controllers and consisted of 10 vol % C H in He (Linde

controlled pneumatic valve. The total volume of C D₆

6

administered to the cell was 0.33 mmol. The sample was

exposed to C D for 10 s followed by evacuation for 1 h. Then

3

8

6

gas) at a total ﬂow rate of 100 mL/min. The WHSV space

a next spectrum of the partially exchanged sample was

recorded. This sequence was automatically repeated to collect

spectra for partially exchanged samples with exposure times of

−1

velocity was kept at 11.7 h . The product composition was

analyzed by an online three-column gas chromatograph

30 s, 5 min, 10 min, 20 min, and 30 min at 30 °C, 30 and 60

(Compact GC Interscience) equipped with a PLOT Al O /

2 3

min at 50 °C, 30 min at 100 °C. Diﬀerence spectra were

obtained by subtracting the initial spectrum of the dehydrated

sample from the spectra after exposure to C D .

Infrared Spectroscopy of Adsorbed CO. Infrared spectros-

copy measurements of adsorbed CO were performed at a liquid

nitrogen temperature using a controlled atmosphere trans-

mission cell cooled by ﬂowing liquid nitrogen through a

capillary spiraled around the catalyst wafer. After the liquid

KCl column with a ﬂame ionization detector and Molsieve-5 Å

and RTX-1 columns both employing thermal conductivity

detectors. The conversion was kept below 2% to ensure

diﬀerential conditions. The reaction rates (r) were calculated

according to r =(X·F)/mcat, where X is conversion of propane, F

6

−1

is ﬂow rate in mol s , and m is the weight of the catalyst in g.

cat

Turnover frequencies were computed from this rate, and the

Brønsted acid site density was determined by H/D exchange.

1

506

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

. RESULTS

.1. TMA CVD on AHFSY Zeolite. In Situ FTIR

Characterization. The reaction of TMA with zeolitic BAS of

dehydrated AHFSY was monitored by in situ FTIR spectros-

copy. Figure 1A shows the spectral changes in the OH and CH

Research Article

3

After TMA adsorption, the spectrum contains a set of bands

in the region 2800−3000 cm due to C−H vibrations

spectrum b in Figure 1 A and B). A small amount of BAS

remains visible after prolonged exposure to TMA, which may

be due to steric hindrances imposed by the cationic complexes.

−1

3

(

Upon reduction in static H (AHFSY-TMA-Red, 100−400 °C)

2

followed by evacuation (spectra c−g, Figure 1 A) bands at 2940,

898, 2865, and 2830 cm⁻corresponding to symmetric and

1

2

asymmetric C−H stretching of CH gradually decrease in

3

intensity. Treatment with H₂at 500 °C is required to

completely eliminate these bands. Concomitantly, it is seen

that a small fraction of the original zeolitic BAS is regenerated.

This regeneration becomes more prominent with increasing

temperature. The nature of the charge-compensating Al species

formed during the reaction with H is not clear. As in principle

2

there is no external source of oxygen, it may on the one hand

be suggested that reduction of the grafted organo-aluminum

intermediates should result in aluminum hydride (Al−H)

species, whose formation may indeed be surmised by

observation of a very weak band around 1932 cm⁻

1

(

Supporting Information, Figure S1). The intensity of this

band is the highest after exposing the AHFSY-TMA to H at

2

1

00 °C. It gradually decreases following reduction at higher

temperatures, and it is completely eliminated after treatment in

H at 500 °C. On the other hand, as will be shown below by

2

XRD and NMR, the reductive treatment leads also to the

decomposition of the zeolite framework, abstraction of

framework oxygen ions, and formation of strong Si-CH₃

bonds, which can provide the oxygen species that are used to

form extraframework Al-oxo complexes. Indeed, a weak band at

3778 cm⁻observed in the spectra of activated catalysts can be

1

Figure 1. In situ FTIR spectra of the CVD of TMA onto AHFSY

zeolite and the decomposition of the grafted organo-aluminum species

via reduction (A) and oxidation (B). (a) Room temperature spectrum

of the dehydrated parent AHFSY, (b) exposed to TMA, (c) evacuated

for 20 min at 50 °C, followed by reduction in H (A) or O (B) (50

64

assigned to hydroxyl groups associated with the EFAl phase.

An alternative procedure for the decomposition of the

grafted cations in O atmosphere (AHFSY-TMA-Oxd) was also

2

investigated. Upon oxidation at 100 and 200 °C (spectra d and e

in Figure 1 B), C−H stretching bands due to grafted TMA

gradually decrease with the concomitant appearance of several

new bands at intermediate temperatures. These bands are due

to oxidation products of Al-CH₃groups and oligomeric

hydrocarbon species. These bands completely disappear after

mbar) and subsequent evacuation at (d) 100 °C, (e) 200 °C, (f) 300

C, (g) 400 °C, (h) 500 °C.

°

stretching regions upon exposure of AHFSY to TMA and

subsequent treatment in H . The introduction of TMA at 50

treatment in O at 500 °C. The oxidation of the methyl groups

2

°

C (spectra a−b in Figure 1) results in a nearly complete

results in carbon oxides and water. In the presence of water, the

cationic Al species will hydrolyze, which will lead to

regeneration of the BAS as well as further agglomeration of

EFAl. Indeed, comparison of the IR spectra after H and O

disappearance of all bands in the hydroxyl region including

−1

those associated with Brønsted acid sites (3549 and 3631 cm )

in sodalite cages and faujasite supercage, respectively, as well as

2

−1

with silanol groups (3744 cm ). This can be tentatively

ascribed to the stoichiometric replacement of Brønsted acid

sites and silanol groups at the external zeolite surface with

treatments evidence more pronounced regeneration of BAS for

the latter treatment.

When the reduction treatment was followed by oxidation in

Al(CH₃)_n⁽

3−n)+

ions.

63

O at 500 °C (AHFSY-TMA-Red-Oxd), the FTIR spectrum

2

a

Table 1. Textural Properties Determined by Ar Physisorption and the Chemical Composition and Unit Cell Dimensions (a )

0

of Diﬀerent Faujasite Catalysts As Well As Framework Al (Al ) Density Derived from XRD

F

c(Al)^I^C^P

(mmol/g)

Al

Si

c(Al_F)^X^R^D

b

S

V

meso

3

BET

(m /g)

micro

(m /g)

meso

(m /g)

micro

(cm /g)

2

3

sample

(wt %) (wt %)

a₀(Å)

(mmol/g)

(cm /g)

c

AHFSY

2.9

3.9

3.0

3.2

3.0

7.8

10.6

8.2

8.6

8.1

n.d.

24.536

24.460

24.369

24.353

24.261

3.25

539

383

391

317

383

468

332

305

256

332

70

51

86

61

51

0.21

0.15

0.13

0.15

0.09

0.12

0.08

0.09

AHFSY-TMA-Hyd

AHFSY-TMA-Red

AHFSY-TMA-Oxd

AHFSY-TMA-Red-Oxd

34.5

30.9

31.4

31.6

2.42

1.436

1.248

0.245

a

S_B_E_T, BET surface area; S

and S_m_e_s_o, microporous and mesoporous surface areas, respectively; V

and V_m_e_s_o, microporous and mesoporous

micro

b

14

pore volumes, respectively. Calculated from Al /U.C. = 107.1(a -24.238) under the assumption of a perfect zeolite framework with unit cell

H Si

F

0

c

Al O (i.e., without absorbed water and Al_E_Fspecies). Not determined.

x 384

x

192−x

1

507

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

not shown) did not appreciably change. Nevertheless, one can

Research Article

(

expect that such a sequential activation procedure may lead to a

more selective transformation of partially reduced and

coordinatively unsaturated species formed upon the high-

temperature reduction of AHFSY-TMA-Red into more uniform

AlO structures than those formed upon the direct oxidative

x

decomposition of AHFSY-TMA.

Chemical Composition, Crystallinity, Surface Area, and

Porosity. The X-ray powder diﬀraction patterns of as-

synthesized AHFSY-TMA, AHFSY-TMA-Red, AHFSY-TMA-

Oxd, and AHFSY-TMA-Red-Oxd are shown in the Supporting

Information, Figure S2 A. The crystallinity of the zeolite after

CVD of TMA (AHFSY-TMA) is close to 100%. It however

strongly decreases after high-temperature activation. After

decomposition of the grafted TMA, the crystallinity of the

modiﬁed AHFSY zeolites is in the range 70−82%.

Figure 2. ¹H NMR Hahn-echo spectra of (a) parent dehydrated

AHFSY zeolite and (b) the material exposed to TMA (AHFSY-TMA)

followed by (c) reduction in H at 550 °C (AHFSY-TMA-Red) or (d)

2

a sequential reduction in H and oxidation in O at 550 °C (AHFSY-

2

TMA-Red-Oxd).

More pronounced are the changes in the lattice constant (a )

0

observed after the modiﬁcation of AHFSY with TMA (Table

of TMA results in the disappearance of almost all zeolitic

hydroxyl groups including the BAS located in the faujasite

sodalite (4.56 ppm) and supercages (3.90 ppm). Also the peak

due to silanol groups (1.90 ppm) disappears almost completely.

At the same time, several new signals appear related to grafted

alkylaluminum -Al(CH ) species (broad signal between −0.3

1

). This parameter can be correlated to the density of

7

framework Al ions in the zeolite. The concentration of Al

F

in AHFSY-TMA as determined by XRD is about 25% lower

than that in the parent AHFSY. Furthermore, the lattice Al

content decreases upon the high-temperature activation (Table

3

2

XRD

63

1

⁾^.^T^h^e^l^o^w^e^s^t^e^s^tⁱ^m^a^t^e^d^l^a^t^tⁱ^c^e^A^l^c^oⁿ^t^eⁿ^t⁽^c⁽^A^lF⁾

)

and −0.8 ppm) and silane O Si(CH ) moieties (0.0−0.2

4−n

3 n

corresponding to the smallest unit cell size is for the catalyst

activated in the reduction−oxidation cycle. Further evidence for

strong dealumination of the faujasite framework upon reaction

with TMA can be observed in the only slight increase of the

ppm, Figure 2 (b)). The latter conﬁrms the substantial

structural modiﬁcations of the zeolite lattice. Subsequent

reduction at 500 °C in H leads to decomposition of a major

2

part of the alkylaluminum groups, whereas a substantial amount

of the silane groups remain present in AHFSY-TMA-Red. The

intensity increase of the signal around 0 ppm suggests that

more silane moieties are formed because of the reaction of

grafted -Al(CH₃)₂fragments with the lattice at elevated

temperature. This is in line with the appearance of the T and

ICP

total Al content in the modiﬁed zeolites (c(Al) , Table 1).

Indeed, an almost complete substitution of BAS upon

modiﬁcation of AHFSY with TMA has been observed by

FTIR spectroscopy (Figure 1). Even if one assumes

substitution of every BAS with a monomeric Al(CH₃)₂⁺ion,

the Al and Si content in AHFSY-TMA should be ∼14 and 33

wt % . While ICP analysis conﬁrms the Si content, the Al

content of the TMA-modiﬁed samples is signiﬁcantly lower.

This implies that the thermochemical activation results in loss

of Al, which appears to be independent whether a reductive or

oxidative treatment is employed. A word of caution is in place

2

9

D sites at ∼ −60 and ∼ −20 ppm, respectively, in the Si

CPMAS spectrum of the reduced sample (Supporting

Information, Figure S3). In addition, a signal at 4.4 ppm

1

appears in the H NMR spectrum of AHFSY-TMA-Red, which

may correspond to either Al−H or strongly polarized BAS. This

peak disappears after oxidation, whereas the signal due to the

supercage BAS at 3.9 ppm does not change upon reduction or

oxidation. Furthermore, high-temperature catalyst activation

leads to a signal representing extraframework Al−OH species

(2.6 ppm). Because formation of this feature is already

as the correlation for estimating c(Al ) from XRD is valid for an

F

intact zeolite framework. It can be expected that the interaction

of TMA with the zeolite framework will not only remove part

of the Al but also induce very strong local distortions in the

F

lattice.

observed after reduction in pure H (in other words in the

2

The Ar physisorption data show that TMA-modiﬁed zeolites

have a signiﬁcantly lower BET surface area (S_B_E_T) than the

parent zeolite (Table 1). This leads to a concomitant decrease

of the micropore volume (V_m_i_c_r_o). The diﬀerence should be due

to the inclusion of extraframework Al species in the zeolite. The

lowest micropore surface area is observed for AHFSY-TMA-

Oxd which may be traced back to the formation of signiﬁcant

amounts of water following oxidation of the methyl ligands in

this case. However, the various treatments do not lead to the

formation of signiﬁcant amounts of mesopores as can be

concluded from comparison of the mesopore surface areas and

volumes. It implies that the structural modiﬁcations only occur

absence of any external source of oxygen), it is taken as an

indication that the oxygen atoms in such extraframework

species originate from the lattice. Oxidation of the reduced

catalyst in O at 500 °C leads to complete decomposition of the

2

grafted Al(CH ) and most Si-CH species. Instead, additional

3

n

3

extraframework Al−OH hydroxyl moieties and silanol groups

are formed (AHFSY-TMA-Red-Ox, Figure 2(d)).

1

27

We further analyzed the results of H-{ Al} TRAPDOR

1

27

NMR measurements. With H-{ Al} TRAPDOR NMR one

65,66

can identify protons located in the vicinity of Al centers.

The spectra for the reduced and oxidized AHFSY-TMA

materials are presented in the Supporting Information, Figure

S4. In the case of AHFSY-TMA-Red, signals of all protons

at the local level.

1

3

.2. NMR Spectroscopy. H MAS NMR. To obtain better

including those from the silane (Si-CH ) moieties are aﬀected

3

insight into the changes in AHFSY resulting from CVD of

TMA, the materials were studied by solid-state magic angle

spinning NMR spectroscopy. Representative H NMR spectra

of the parent AHFSY zeolite and the TMA-modiﬁed catalysts

are shown in Figure 2. In agreement with the FTIR data, CVD

by the Al irradiation. It is important to note that the signal at

2

7

4.4 ppm shows the same behavior with regards to the Al

irradiation as the protons due to Al-bound hydroxyl groups. A

more signiﬁcant TRAPDOR eﬀect is expected for the hydrogen

atom directly bound to Al. This would suggest that this peak

1

508

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

should be assigned to strongly polarized BAS rather than to

aluminum hydride species. In principle, for speciﬁc geometries

2

7

1

27

(

regarding the angle between the Al- H dipole and Al

quadrupole tensors) the TRAPDOR eﬀect would also be weak.

However, such speciﬁc geometries would be highly coinciden-

tal. Furthermore, the absence of aluminum hydride species is in

line with the results of FTIR measurements showing that,

although being formed upon hydrogenolysis of grafted TMA,

Al−H species are not present in the ﬁnal reduced zeolite. The

complete disappearance of this absorption band is observed

after reduction at 500 °C (Supporting Information, Figure S1).

1

27

After oxidation the behavior of H signals with respect to Al

irradiation is substantially changed (Supporting Information,

Figure S4). For AHFSY-TMA-Red-Oxd, two diﬀerent Si-CH₃

groups can be distinguished characterized by signals at 0.20

2

7

ppm, which is sensitive to Al irradiation, and at 0 ppm, which

Figure 3. Deconvoluted direct-excitation ²⁹Si MAS NMR spectra of

AHFSY, AHFSY-TMA, AHFSY-TMA-Red, and AHFSY-TMA-Red-

Oxd. The relative intensities of diﬀerent structure units and the

calculated lattice Si/Al ratios are given in Table 2.

27

is unaﬀected by Al irradiation. This suggests that these signals

correspond to the Si-CH groups located in distinctly diﬀerent

chemical environments, most likely Al- and Si-rich regions in

the zeolite. The signal due to SiOH groups is hardly aﬀected by

the ²⁷Al polarization, suggesting that the regenerated silanol

groups are located in Al-poor regions.

3

distortions induced by the sequential reduction−oxidation

2

9

treatment of the TMA-containing AHFSY.

Si MAS NMR. We also investigated how the TMA

29

_m_o_d_i_ﬁ_c_a_t_i_o_n_i_n_ﬂ_u_e_n_c_e_s_t_h_e_s_t_a_t_e_o_f_l_a_t_t_i_c_e_s_i_l_i_c_o_n_a_t_o_m_s_.2 Si

9

The overall Si NMR intensity can be greatly enhanced by

1

29

using H- Si cross-polarization (CP) (Supporting Information,

Figures S3B,C), but the resulting NMR spectra are generally

not quantitative, because proton-rich species tend to be

relatively overemphasized. Interestingly, in addition to the

MAS NMR provides detailed information about the nature and

coordination of the Si atoms. In particular, it allows to estimate

the framework Si/Al ratio from the relative intensities of the

F

2

9

Si NMR signals associated with Si atoms having diﬀerent

4

major Q (nAl) signals already mentioned above (Figure 3,

numbers of Al neighbors Q (nAl), where n stands for the

4

5,67

Supporting Information, Figure S3A), the CP-enhanced

spectrum of AHFSY-TMA-Red shows two broad signals

around −60 ppm and −20 ppm, which are attributed to T-

sites (-SiCH ) and D-sites (-Si(CH ) ), respectively, (Support-

number of Al neighbors.

The direct-excitation (DE) ²9_S_i_M_A_S_N_M_R_s_p_e_c_t_r_u_m_o_f

AHFSY is characterized by three signals of the zeolite

framework corresponding to silicon atoms surrounded by

diﬀerent number of Al atoms, namely Si(2Al), Si(1Al), and

Si(0Al) at chemical shift positions of −95, −101, and −106

ppm, respectively (Supporting Information, Figure S3). The

reaction with TMA (AHFSY-TMA) leads to a strong decrease

3

3 2

ing Information, Figure S3C) formed during reduction. High-

temperature oxidation results in the decomposition of these

silane moiety, as evidenced by the disappearance of the T and

D sites.

27_A_l_M_A_S_a_n_d_M_Q_M_A_S_N_M_R_._A_l_u_m_i_n_u_m_a_t_o_m_s_i_n_p_e_r_f_e_c_t

4

of the signals due to Q (2Al) and Q (1Al), whereas the signal

attributed to Q (0Al) increases. This points to dealumination

crystalline zeolite lattices such as in the parent AHFSY material

are tetrahedrally coordinated and characterized by an isotropic

4

of the zeolite framework. In addition, a new signal at −112 ppm

occurs indicating the formation of regions with a high silica

content. The DE spectra of the reduced and the subsequently

oxidized TMA-modiﬁed AHFSY diﬀer only slightly in the

intensities of Si(nAl) signals. The intensities of the diﬀerent

spectral components were obtained by deconvolution of the

DE ²⁹Si MAS NMR spectra in terms of Gaussian line shapes

using DMﬁt2011. The spectral deconvolution is shown in

Figure 3, and the relative intensities I of the diﬀerent Q (nAl)

signals summarized in Table 2, as well as the Si/Al values

27

Al chemical shift ranging from 55 to 65 ppm (see Supporting

Information, Figure S5). The introduction of extraframework

species into the zeolite pores through chemical treatment or via

27

dealumination of the lattice leads to more complex Al MAS

NMR spectra. The signal of tetrahedral Al becomes broader as

a result of distortions in the oxygen-coordination symmetry

27

around the Al atoms. Furthermore, the resolution and the Al

6

8

NMR visibility suﬀer from the second-order quadrupolar

broadening, which is not completely eliminated by magic-

angle spinning. In addition, the formation of octahedral and

4

n

F

4

5,67,69

calculated by use of the formula Si/Al = 4 (ΣI )/(Σn I ).

In principle, the intensities I and I of Q (1Al) and Q (2Al) Si

EF

F

n

penta-coordinated Al leads to new signals at ∼0 ppm and

4

27

1

2

∼35 ppm, respectively, in Al NMR spectra.

As generally observed for dehydrated zeolites, the ²⁷Al

NMR spectra of the as prepared TMA-modiﬁed zeolites

samples are poorly resolved. They all contain an asymmetric

broad peak at ∼58 ppm (Figure 4D and Supporting

Information, Figure S5A), which is attributed to tetrahedral

70

atoms at −101 and −95 ppm may also comprise contributions

3

from silanol moieties Q (0Al) and Q (1Al) in the TMA-

modiﬁed zeolite lattice. However, the gravimetric Al densities

derived from the calculated Si/Al ratios agree generally well

with the Al concentrations estimated from XRD (Table 4).

Thus, despite the expected damage to the zeolite lattice in

TMA modiﬁed AHFSY, the silanol concentration is still

relatively low compared to the density of Si atoms with Al

neighbors. An obvious exception is the AHFSY-TMA-Red-Oxd

case, where the value derived from the XRD lattice constant is

much lower. This is probably caused by the strong lattice

F

IV

aluminum (Al_F). For AHFSY-TMA, the spectrum contains a

number of overlapping spinning sideband patterns in the range

from 200 ppm to −200 ppm, indicative of the heterogeneous Al

speciation and strong quadrupolar interactions of framework

and extraframework Al with distorted oxygen-coordination

symmetry. Upon reduction (AHFSY-TMA-Red), the spectrum

1

509

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

2

9

Table 2. Relative Intensities of the Si NMR Signal Components and Calculated Si/Al Ratios for AHFSY and the As Prepared

F

(Dehydrated) TMA-Modiﬁed Zeolites

Si(2Al) (%)

Si(1Al) (%)

Si(0Al) (%)

samples

−95 ppm

−101 ppm

−107 ppm

−112 ppm

Si/Al (NMR)

F

AHFSY

21

10

7

55

37

33

23

24

51

53

4.1

6.9

8.4

9.0

AHFSY-TMA

2

9

AHFSY-TMA-Red

AHFSY-TMA-Red-Oxd

11

13

Figure 4. ²⁷Al MAS NMR spectra of the TMA-modiﬁed zeolites. (A,B) 1D single-pulse-excitation spectra of air-exposed, hydrated samples analyzed

in diﬀerent ways (see below); (C,D) 2D MQMAS NMR spectra of (C) air-exposed hydrated zeolites and (D) the same materials in the dehydrated

state before exposure to moist air. Projections (P) on the (horizontal) direct and (vertical) indirect frequency axis are indicated along the respective

2

D spectra. Above the 2D spectra of the dehydrated samples in (D) the respective 1D spectra are also shown. All spectra were analyzed in terms of

line shape components involving Gaussian distributions of isotropic chemical shifts and quadrupolar coupling constants. The line shape parameters

resulting from spectral ﬁtting are given in Table 3. For the line shape deconvolution in (B) the same model as in (C) is used with only the signal-

component heights as free ﬁtting parameters. The 1D line shape components predicted on the basis of the simulated MQMAS spectra are too

narrow. The analysis of the 1D spectra in (A) is based on a less restricted line shape ﬁt with adjustable center C_Qand the width ΔC_Qof the

quadrupolar coupling constant distribution (Table 3).

contains two overlapping resonances, one around 55 ppm and

the other one around 35 ppm. The peaks broaden further upon

oxidative treatment, suggesting an increase in the heterogeneity

of the Al species.

To get more detailed insight into the nature of Al species in

modiﬁed zeolites, the multiple-quantum magic angle spinning

second-order quadrupolar eﬀect is reﬂected by broadening

along the (horizontal) direct-frequency axis. In addition there is

also a quadrupolar-coupling eﬀect on the signal positions. The

stronger the quadrupolar coupling, the further the resonances

67

are separated from the spectral diagonal. MQMAS spectra of

the as-prepared, dehydrated TMA-modiﬁed zeolites show

indeed increased spectral resolution (Figure 4D). Two

resonances with isotropic shifts ∼58 and ∼30 ppm can already

be recognized by visual inspection of the spectra of AHFSY-

TMA-Red and AHFSY-TMA-Red-Oxd. The 58 ppm signal is

attributed to tetrahedral framework Al, and the 30 ppm signal

to penta-coordinated extraframework Al. Detailed line shape

simulation by use of “Czjzek” line shape components based on

a Gaussian distribution of isotropic shifts and quadrupolar

(MQMAS) NMR spectra were recorded. This technique oﬀers

the possibility to separate the quadrupolar induced shift and

broadening of the ²⁷Al resonances reﬂecting local electric ﬁeld

gradients around the Al atoms, on the one hand, and the

isotropic chemical shifts typical of the chemical structure

around the Al atoms, such as the oxygen coordination state, on

the other hand. Chemical heterogeneity shows up as broad-

ening along the diagonal in the 2D spectrum, whereas the

1

510

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

a

27

Table 3. Lineshape Parameters of the Signal Components in Al MAS NMR Spectra of the Parent AHFSY and TMA-Modiﬁed

Zeolites

Al site

T

δ_i_s_o(Δδ ) (ppm)

C_Q(ΔC ) (MHz)

C′_Q(ΔC′ ) (MHz)

fraction

1

c(Al ) mmol/g

iso

Q

F

AHFSY

hydrated

61 (5)

2.1 (1.2)

0.8 (0.5)

2.9

AHFSY-TMA

dehydrated

hydrated

T

P

59 (7)

62 (7)

35 (4)

5 (5)

3.9 (2.0)

2.9 (1.1)

3.6 (1.6)

3.8 (1.8)

3.4 (2.0)

8.0 (5.1)

4.4 (2.6)

0.47

0.15

0.37

1.7

0.5

1.3

O

AHFSY-TMA-Red

dehydrated

T₁

T₂

P

59 (7)

45 (7)

30 (10)

61 (5)

33 (6)

5 (6)

3.9 (2.0)

3.6 (1.6)

3.3 (1.7)

4.2 (2.0)

3.4 (1.8)

hydrated

T

4.7 (2.8)

5.2 (2.9)

4.5 (2.7)

0.48

0.09

0.43

1.7

0.3

1.5

P

O

AHFSY-TMA-Oxd

hydrated

T

P

63 (7)

36 (8)

7 (6)

3.8 (1.1)

4.8 (1.6)

4.0 (1.6)

5.0 (2.9)

8.4 (5.6)

5.0 (2.5)

0.41

0.19

0.

1.3

0.6

1.2

O

AHFSY-TMA-Red-Oxd

dehydrated

T₁

T₂

P

56 (7)

46 (7)

30 (10)

60 (8)

33 (7)

4 (6)

4.9 (1.8)

4.0 (1.3)

4.8 (1.8)

3.5 (1.6)

4.0 (1.8)

3.2 (1.7)

hydrated

T

4.7 (4.1)

5.1 (2.2)

4.7 (3.1)

0.49

0.06

0.45

1.4

0.2

1.3

P

O

a

δ , average isotropic chemical shift; Δδ , full-width-at-half-maximum (fwhm) of the Gaussian δ distribution; C , average quadrupolar coupling

iso

Q

constant; ΔC , fwhm of the Gaussian C distribution. C and ΔC determined from MQMAS spectra; C′ , ΔC′ and signal-intensity fractions

determined from 1D Al NMR spectra of the hydrated zeolites; Al , Al and Al densities from the intensity fractions and the total Al content

c(Al)_t_o_t

Q

27

IV

V

VI

Al NMR

in Table 3. T, tetrahedral Al; P, pentacoordinated Al; O, octahedral Al.

6

8,71

1D and 2D ²⁷Al NMR spectra of the hydrated TMA-

modiﬁed samples indeed show much better resolved signals

attributed octahedral, penta-coordinated, and tetrahedral Al

atoms at ∼5, ∼33, and ∼61 ppm, respectively (Figures, 4A−C).

The relative amount of octahedral Al increases upon hydrogen

and oxygen treatment. This trend correlates with the

appearance of penta-coordinated Al in as-prepared dehydrated

AHFSY-TMA-Red and AHFSY-TMA-Red-Oxd (1D NMR

spectra in Figure 4D). This suggests that hydration causes

conversion from originally penta-coordinated Al into octahedral

Al species. Water molecules probably bind to under-

coordinated Al centers. The absolute intensity of the

tetrahedral Al signal of the parent AHFSY zeolite is lower

than those of the TMA-modiﬁed zeolites (Supporting

Information, Figure S5B). This suggests that grafting of TMA

leads to dealumination of the faujasite framework, which would

also be in line with the breakage of the zeolite lattice reﬂected

coupling constants

reveals that there is a third signal

around 45 ppm (Table 3). A signal at this position has

previously be assigned to tetrahedral extraframework Al.

72

The projections of the MQMAS spectra on the single-

quantum frequency axis and the corresponding 1D single-pulse-

excitation spectra are strongly diﬀerent (indicated above the 2D

spectra in Figure 4D). As can be observed, only the (narrower)

signals of ²⁷Al spins with moderate quadrupolar coupling

interactions show up in the MQMAS spectra recorded at 11.7

T. This is caused by the reduced triple-coherence generation

for Al spins with quadrupolar coupling stronger than the radio

frequency ﬁelds employed for excitation. The MQMAS spectra

thus reﬂect only part of the heterogeneous Al sites. One way to

make MQMAS less selective and be able to detect (almost) all

Al spins is to reduce the strong quadrupolar interactions by

hydration of the zeolites. The presence of water molecules with

their electrical dipoles decreases the local electronic ﬁeld

gradients around the Al nuclei. This results in weaker

quadrupolar coupling interactions, causing less quadrupolar

27

2

7

29

in Si NMR spectra (formation of D and T type silicon).

2

7

VI

Therefore, the octahedral (Al ) and pentacoordinated

EF

V

(

Al_E_F) Al species may not be exclusively related to the

27

broadening and thus enhancing the chemical resolution in Al

MAS NMR. Unfortunately, for the TMA-modiﬁed AHFSY

zeolites it will also signiﬁcantly alter the chemical environment

of the Al centers. Nevertheless, if the eﬀect of water and air is

additional Al atoms from TMA, but may also arise from the

partial framework dealumination and/or severe structural

IV

distortions of the originally tetrahedral lattice Al_Fsites.

Subsequent high-temperature reduction and oxidation treat-

ments lead to a further decrease of the signal at 62 ppm

belonging to tetrahedral Al sites and an increase of those

attributed to more coordinatively saturated extraframework

species. In the chemical vapor deposition procedure the initial

TMA dose is much higher than the available Brønsted sites for

2

7

taken into account, Al NMR of the hydrated TMA-modiﬁed

zeolites may still give “retrospective” insight into structures

formed and the lattice damage resulting from TMA deposition

and consequent high-temperature treatments with hydrogen

and oxygen.

1

511

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

Table 4. Total Al Concentrations from ICP (c(Al) ^I^C^P) and ²⁷Al NMR (c(Al)_t_o_t^A^l^N^M^R), and Framework Al Concentration from

tot

27

Al NMR 29

Si NMR

XRD

Al NMR (c(Al_F)

), Si NMR (c(Al_F)

), and XRD (c(Al_F)

)

Sample

c(Al)_t_o_t^I^C^P(mmol/g)

c(Al)_t_o_t^A^l^N^M^R(mmol/g)^a

c(Al_F)^A^l^N^M^R(mmol/g)^b

c(Al_F)^Sⁱ^N^M^R(mmol/g)^c

c(Al_F)^X^R^D(mmol/g)^d

AHFSY

2.9

3.9

3.0

3.2

3.0

2.9

3.5

3.1

2.9

1.7

1.3

1.4

3.2

2.3

1.5

1.9

3.25

2.42

1.44

1.25

0.46

AHFSY-TMA

AHFSY-TMA-Red

AHFSY-TMA-Oxd

AHFSY-TMA-Red-Oxd

a

From total ²⁷Al NMR peak area [120, −80 ppm] of hydrated TMA-modiﬁed AHFSY zeolites relative to AHFSY with 2.9 mmol/g Al (ICP value).

b

IV

Al NMR

c

From relative intensity of the Al signal of hydrated samples (Table 3) combined with c(Al)

. Framework Al density from the Si/Al ratio

tot

derived from ²⁹Si NMR (Table 2) combined with 16.7 mmol T atoms/g zeolite framework. c(Al_F)

d

XRD

=10.79 (a -24.238) with a the lattice

0

constant (Å) assuming a unit cell H Si Al O₃₈₄.

x

192‑x

x

Figure 5. FTIR spectra of dehydrated AHFSY-TMA-Red (A) and AHFSY-TMA-Oxd (B) zeolites prepared in the regions of OH (left) and OD

(

right) stretching vibrations before and after reaction with C D with increasing exposure times at 30 °C followed by the gradual increase of the

6 6

reaction temperature to 100 °C.

charge compensation. So, that it is likely that a fraction of EFAl

exists as oligomeric oxygenated Al clusters or even aggregated

aluminum oxide species.

analysis. This resulted in a quite reasonable ﬁt to the 1D

spectra, but the line shape components extracted from the

MQMAS spectra are too narrow for the 1D spectra (Figure

4B). Apparently, even after hydration, part of the Al with

stronger quadrupolar couplings is under-represented in

MQMAS. A better spectral ﬁt results if the Gaussian-

distribution center and width of the quadrupolar coupling

constants C and ΔC is allowed to change. In general, the

2

7

To quantitatively analyze the 1D Al NMR spectra of the

IV

V

VI

hydrated zeolites in terms of Al , Al , and Al a proper line

shape model is required, which is not trivial because of the

spectral overlap and the asymmetric character of the second-

order quadrupolar linebroadening. Therefore, the individual

line shape parameters were ﬁrst determined from MQMAS

NMR spectra, because of the lower signal overlap in the 2D

spectra, and because of the additional information content in

the exact shift position (separation from the spectrum diagonal)

and line shape (horizontal/diagonal line broadening) of the

resonances. Deconvolution of these spectra in terms of the

earlier mentioned Czjzek components are summarized in Table

Q

overall C and ΔC values increase compared to those in the

Q

MQMAS spectra.

If ²⁷Al NMR visibility in the hydrated parent and TMA-

modiﬁed zeolites is similar, the (weight-normalized) peak area

2

7

of the central-transition frequency region in the respective Al

NMR spectra should be proportional to the total Al

ICP

concentration measured by ICP, c(Al)_t_o_t_a_l, which is roughly

3

. However, MQMAS spectra are not quantitative in the sense

the case (Table 4). The slightly lower values from NMR for

hydrated AHFSY-TMA, AHFSY-TMA-Oxd, and AHFSY-Red-

Oxd indicate that a small Al fraction is NMR invisible as a result

of low coordination symmetry. From the total Al content and

the relative signal intensities for the Al , Al , and Al species

obtained from the deconvolution of the 1D NMR spectra

(Figure 4A), the individual content of the diﬀerent Al species

that the signal intensities do not directly reﬂect the relative

occurrence of the various Al types in the material. Therefore,

the MQMAS models were ﬁtted with minimal changes to the

IV

V

VI

1

D single-pulse excitation spectra, which are quantitative, but

suﬀer from higher signal overlap. First, all line shape

parameters, except the signal heights, were kept ﬁxed in the

1

512

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

can be calculated (Table 3). The Al densities from Al NMR,

Research Article

IV

27

Al NMR

c(Al )

, correlate reasonably with the Al densities from

F

2

9

Si NMR

XRD

Si NMR (c(Al_F)

) and XRD (c(Al_F) ) (Table 4), but

are systematically lower. This is the case, even for AHFSY,

where ²⁷Al NMR visibility should not be an issue. A likely

explanation is that c(Al_F)^A^l^N^M^Ris indirectly calibrated with

respect to the ICP value for hydrated AHFSY, thus including

Si NMR

XRD

adsorbed water. In contrast, c(Al_F)

and c(Al_F)

are

speciﬁed per gram zeolite lattice framework, excluding

extraframework species.

3

.3. Brønsted and Lewis Acidity. H/D Exchange with

Perdeuterated Benzene. Selective H/D exchange of perdeu-

terated benzene with zeolitic BAS was monitored by FTIR

spectroscopy to investigate the eﬀect of TMA-modiﬁcation on

the intrinsic acidity of AHFSY zeolites. After the decomposition

of grafted TMA under reducing condition at high temperature

and prior to reaction with C D , the OH stretching region of

Figure 6. First-order plot of the rate of appearance of the band at

about 2680 cm due to acidic deuteroxyl groups in the faujasite

supercage upon reacting the parent AHFSY and TMA-modiﬁed

−1

zeolites with C D . c is the concentration of the respective OD

6

(HF)OD

6

groups at a particular stage of the H/D exchange at 30 °C, N₍_H_F₎_O_Dis

the total density of acid sites, t′exchange is the exposure time to C

TMA-AHFSY-Red zeolite contains two absorption bands with

−1

D .

6 6

maxima at 3549 and 3631 cm corresponding to the BAS

located in faujasite sodalite cages and supercages, respectively

modiﬁed zeolites are much lower than those of EFAl-containing

zeolites prepared by ion exchange but similar to the EFAl-

containing zeolites prepared via incipient wetness impregna-

(

Figure 5). Upon reaction with C D , the intensity of these

6 6

bands gradually decreases, while new bands appear in the OD

stretching region (Figure 5 A, right panel). Most of the BAS in

AHFSY-TMA-Red react with C D already after prolonged

35

tion. The decreased intrinsic acidity of TMA-modiﬁed

6

zeolites is surprising taking into account the pronounced

dealumination of the zeolite framework in these materials (see

section 3.2).

contact at 30 °C. Deuteration is nearly complete after reaction

at 100 °C. The total acidity of AHFSY-TMA-Red is 0.3 mmol/

g (Table 5), amounting to 10% of the total number of acid sites

in AHFSY.

Infrared Spectroscopy of Adsorbed CO. The presence and

strength of the Lewis and Brønsted acid sites in TMA-modiﬁed

faujasites was further probed by FTIR spectroscopy of CO

adsorption at liquid nitrogen temperature. Both the perturba-

tions of adsorbed CO and those of the zeolitic hydroxyl groups

due to hydrogen bonding with CO molecules were monitored.

The FTIR spectra in the region of OH and CO stretching

vibrations (ν(OH) and ν(CO), respectively) at increasing CO

coverage for the TMA-modiﬁed zeolites are collected in the

Supporting Information, Figure S6. The red shifts of ν(OH)

Table 5. Initial C D H/D Exchange Rates (k_H_/_D) at 303 K

6

FTIR

and Total Acidity (c(BAS) ) of Diﬀerent Faujasite

Catalysts

k_H_/_D₍_h₋1

_c₍_B_A_S₎FTIR (mmol/g)

a

sample

)

AHFSY

1.1

0.3

0.4

4.7

3.3

0.3

0.7

1.3

AHFSY-TMA-Red

AHFSY-TMA-Oxd

AHFSY-TMA-Red-Oxd

USY-8

(

denoted as Δν(OH)) are summarized in Table 6. The spectra

a

Table 6. Positions of ν(CO) and ν(OH) Stretching Bands

for OH···CO Complexes Formed after Adsorption of CO on

Total concentration of BAS determined by the H/D exchange of

C D .

6

a

Zeolite Catalysts at the Liquid Nitrogen Temperature

The reaction of the oxygen-treated AHFSY-TMA zeolite

ν(CO)

ν(OH)

(cm

3338

Δν(OH)

−1

catalyst

(cm

)

(cm

294

)

with C D is much slower (Figure 5 B). In this case only a small

6

fraction of BAS can be exchanged at 30 °C. The H/D exchange

remains slow at a temperature of 50 °C. Despite an apparent

lower intrinsic acidity, the concentration of BAS in AHFSY-

TMA-Oxd (0.76 mmol/g) is approximately twice higher than

in its reduced counterpart. The exchange behavior and the total

number of exchangeable BAS in the AHFSY-Red-Oxd are very

similar to that in AHFSY-TMA-Oxd (Table 5).

The intrinsic acidity of the modiﬁed zeolites can be

compared by analyzing the rate of appearance of OD stretching

vibrations in the FTIR spectra (Figure 6). The ﬁrst-order rate

constants are given in Table 5. The initial H/D exchange rate

AHFSY

2177

2176

AHFSY-TMA-Red

AHFSY-TMA-Oxd

AHFSY-TMA-Red-Oxd

3329

3332

3329

300

301

300

a

Shifts are with respect to the vibration of the hydroxyl groups in the

supercages.

−1

of adsorbed CO are characterized by ν(CO) = 2177 cm . Both

ν(CO) and Δν(OH) are very similar for all zeolites. The red

shift is slightly higher for the modiﬁed zeolites than for the

EFAl-free AHFSY (Table 6).

for the TMA-modiﬁed AHFSY zeolite activated in H (AHFSY-

3.4. Catalytic Reactivity. The catalytc activity data

2

TMA-Red) is similar to that of the original EFAl-free AHFSY

zeolite. At higher exchange levels, the rate decreases. This may

point to some heterogeneity in the strength of the BAS.

However, the samples activated by high-temperature oxidation

including the overall reaction rate (r), the rates of propane

DEH

dehydrogenation to H and propylene (r ) and of propane

2

CRC

cracking to ethylene and methane (r ) as well as turnover

frequencies are collected in Table 7. In all cases, the catalytic

activity was stable during the catalytic runs (Supporting

Information, Figure S7). The overall conversion rate of the

TMA-modiﬁed zeolites depends strongly on the pretreatment.

(

AHFSY-TMA-Oxd and AHFSY-TMA-Red-Oxd) are clearly

substantially less active in the H/D exchange reaction (Figure

, Table 5). Noteworthy is also that the rates for the TMA-

6

1

513

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

a

Table 7. Catalytic Conversion of Propane over TMA-Modiﬁed AHFSY Catalysts and Commercial USY Zeolite

c(BAS)^F^T^I^R

r × 10

−5

r^C^R^C× 10⁻⁵

(mol/g·s)

r^D^E^H× 10⁻⁵

(mol/g·s)

CRC

(%)

TOF

CRC

(10⁻³

TOF

DEH

(10⁻³

sample

(mmol/g)

(mol/g·s)

s⁻¹)

s⁻¹

7.8

)

AHFSY

3.34

0.34

0.73

1.29

14.8

29

17

9

11.8

14

2.6

15

82

48

41

66

60

35

412

95

AHFSY-TMA-Red

AHFSY-TMA-Oxd

AHFSY-TMA-Red-Oxd

441

134

42

6.9

5.9

13

9.8

3.1

8.5

80

USY-8

21

101

66

a

−1

WHSV = 11.7 h ; T = 590 °C.

The rate for AHFSY-TMA-Red is twice as high as the rate for

the parent zeolite. It is even higher than the activity of the

commercial steam-calcined USY-8 zeolite. After oxidation of

this sample (AHFSY-TMA-Red-Oxd) the activity is lower than

the parent zeolite. The activity of AHFSY-TMA-Oxd is close to

that of AHFSY. The carbon-containing products from propane

conversion under these conditions are methane, ethylene, and

propylene. For the parent zeolite AHFSY, the rate of propane

cracking with methane and ethylene as main products is much

higher than the rate of propane dehydrogenation. The

contribution of cracking to propane conversion (CRC) is

oxygen or hydrogen atmosphere at elevated temperature. The

IR spectra of TMA-modiﬁed zeolites diﬀer substantially with

respect to the way the grafted organoaluminum species are

treated. The spectra contain indications that their hydro-

genolysis in H results in the formation of Al−H species

2

−

1

characterized by a weak band at around 1932 cm at

intermediate reduction temperature. At the highest temper-

ature, these species and all methyl groups have been reacted

away. These changes go together with partial regeneration of

the zeolitic BAS in AHFSY-TMA-Red. The amount of Al

deposited during CVD of TMA (cf. AHFSY-TMA-Hyd in

Table 1) corresponds to approximately 1/3 of the number of

BAS initially present. This suggests that the organometallic

precursor undergoes multiple reactions with the hydroxyl

groups of the zeolite. As the Al content of AHFSY-TMA-Red is

found to be lower than that of the hydrated zeolite, it can be

inferred that part of the deposited Al species has been removed

during reduction. Combined with the observation of the

transient formation of Al-hydride species, it may be speculated

that during the high-temperature treatment volatile (AlH ) has

82%. Upon modiﬁcation by TMA and reduction, the rate of

propane dehydrogenation increases nearly 6 times, while the

rate of cracking only increases slightly. As a result, the

contribution of propane cracking for this zeolite has decreased

to 48%. For AHSY-TMA-Oxd, this contribution is even lower,

which is due to the decreased (increased) rate of cracking

(

dehydrogenation) compared to the parent zeolite. Note that

the increase in dehydrogenation is less than for AHFSY-TMA-

Red. When the reduced zeolite is oxidized (AHFSY-TMA-Red-

Oxd), the rates of propane cracking and dehydrogenation

decrease strongly, the latter more pronouncedly than the

former so that the contribution of cracking increases to 66%.

Notably, the rate of propane cracking over USY-8 is almost as

high as that of AHFSY-TMA-Red, while its dehydrogenation

activity is close to that of AHFSY-TMA-Oxd. Except for

AHFSY-TMA-Red, in all cases the molar methane-to-ethylene

ratio was very close to unity. It is 1.29 for AHFSY-TMA-Red

indicating that the ethylene product is further converted.

Indeed, this is the only zeolite for which the formation of a

small amount of benzene was observed. The benzene selectivity

was, however, below 1%. This may suggest that also some coke

formation takes place which did not lead to catalyst

deactivation at the time scale of the reaction test.

3

n

been formed. The existence of such compounds has recently

73

been reported. In support of this, it was observed that

downstream the zeolite bed in the quartz reactor used for in

situ reduction a white powder was formed. This is likely due to

the reaction of volatile aluminum hydride species with hydroxyl

groups of the surface of the quartz reactor. The oxidative

decomposition of the grafted organoaluminum species

(

AHFSY-TMA-Oxd) results in water and carbon oxides. In

the presence of water the Al−O bonds are hydrolyzed, as

follows from the much more pronounced regeneration of

zeolitic BAS following oxidation compared to reduction.

Notably, in this case the Al content also decreases during the

treatment.

The intensive reaction of TMA with more than one BAS

results in signiﬁcant distortions of the zeolite lattice as follows

from XRD and Ar physisorption measurements. MAS NMR

conﬁrms the partial decomposition and dealumination of the

Table 7 also contains the turnover frequencies (TOF) for

propane cracking and dehydrogenation. These TOF values

were computed on the basis of the BAS densities determined

by the H/D exchange method. The TOF for propane cracking

of AHFSY-TMA-Red is an order of magnitude higher than that

of the parent zeolite and about 4 times higher than that of USY-

1

framework. H NMR shows that during CVD of TMA nearly

all zeolite hydroxyl groups are consumed. Concomitantly, Al-

CH and Si-CH species are formed. Thus, it can be inferred

3

8

. The TOF for propane dehydrogenation is also much higher

that CVD does not proceed by simple reactions as encountered

for AFHSY-TMA-Red than for the two reference cases. The

TOFs for cracking and dehydrogenation of the oxidized TMA-

modiﬁed zeolites are substantially lower than the respective

values for their reduced counterpart.

39

earlier for CVD of Zn(CH ) onto high-silica ZSM-5 zeolite,

3

2

+

involving formation of Zn(CH₃) and methane. Besides

reaction with the BAS, TMA will also attack Al−O−Si

framework moieties. As Si-CH bonds are stronger than Al-

3

CH bonds, it is likely that the rupture of the relatively weak

framework bonds around the bridging hydroxyl group leads to

3

4

. DISCUSSION

The in situ FTIR data show that BAS of dealuminated zeolite Y

can be made to react with TMA. Grafting of TMA onto the

zeolite leads to complete removal of all hydroxyls, including the

silanol and bridging hydroxyl groups. The resulting organo-

metallic intermediates can be decomposed by treatment in

Si-CH and Al−O−{TMA} species. This is consistent with

3

NMR and IR spectroscopic data. High-temperature treatment

of AHFSY-TMA in H results in the decomposition of the Al-

2

CH species, while the Si-CH moieties remain intact. The

3

latter are only decomposed when oxidized at 500 °C in O₂

1

514

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

resulting in silanol groups. After reduction as well as after

subsequent oxidation, a small portion of BAS is regenerated.

Their regeneration is however stronger during the direct

oxidative treatment. The nature of the regenerated protons

depends on the activation method. The protons in AHFSY-

TMA-Red give rise to an unusual resonance at 4.4 ppm in the

Despite these relatively small diﬀerences in the intrinsic

Brønsted acidity, the TMA modiﬁcation has a signiﬁcant eﬀect

on the rate of propane conversion. From comparison of weight-

based activities, one concludes that AHFSY-TMA-Red is more

active than the parent AHFSY as well as a commercial USY

cracking zeolite. Besides, when the activities are normalized per

BAS, all of the TMA-modiﬁed zeolites exhibit a higher intrinsic

propane conversion rate than the two reference zeolites.

Importantly, the modiﬁcation with TMA leads to a signiﬁcant

increase in the rate of propane dehydrogenation. The results

support the previous proposal that Lewis acidic Al sites are

1

H NMR spectrum. These are tentatively assigned to strongly

polarized protons in the supercages. We propose that the

polarization is caused by the interaction with neighboring Lewis

acidic extraframework Al species. These protons were not

observed anymore after oxidation of the zeolite sample. In such

35

1

involved in propane dehydrogenation. From Table 7 and the

characterization results, it may be inferred then that the rate of

propane dehydrogenation is the highest for the strong Lewis

acidic groups in reduced AHFSY-TMA. Oxidation results in the

formation of a weaker form of Lewis acidity, and the Al

dispersion will likely also be lower. In any case, the oxidized Al

species show a much lower rate of propane dehydrogenation.

The most striking ﬁnding of the present study is the

exceptionally high cracking activity of AHFSY-TMA-Red.

Despite its low BAS density, the overall propane conversion

rate was higher than that of the commercial USY zeolite. When

normalized per BAS, the propane cracking activity is many

times higher than that of the reference zeolites. In our previous

case, H resonances characteristic for faujasite zeolites were

observed again. The decomposition of the intermediate silane

species and the dealumination of the zeolite framework follows

_f_r_o_m_t_h_e2 Si and Al NMR data. The Al NMR data also

9

27

inform us about the fate of Al species. After reductive treatment

of grafted TMA species, the Al species are mainly 4- and 5-

coordinated. Oxidation of these species leads to a more

heterogeneous Al speciation. From the strong quadrupolar

2

7

broadening of the signals in the Al NMR spectra of

dehydrated materials, it can be concluded that a substantial

fraction of Al cannot be detected by NMR spectroscopy.

Although this problem can be partially avoided by hydrating the

zeolite samples, this procedure strongly alters the chemical

environment of Al sites of the TMA-modiﬁed zeolites. The

important corollary of the NMR results appears to be that

dealumination occurs upon CVD of TMA. This is consistent

35

study, we proposed the existence of a correlation between the

intrinsic cracking activity of EFAl-containing zeolites and the

ratio of cationic extraframework aluminum species and BAS.

Figure 7 shows this correlation including the data for the

IV

with the lower concentration of framework Al species derived

from XRD analysis. On the basis of the IR, XRD, and NMR

results, it can be concluded that the TMA-modiﬁed zeolites

contain a signiﬁcant part of charge-compensating Al ions.

The strong structural changes in the zeolites resulting from

the reaction with TMA and subsequent treatment signiﬁcantly

aﬀect their acidic properties. H/D exchange with C D₆

6

measurements do not evidence an enhancement of the acidity

of bridging hydroxyl groups in the modiﬁed zeolites. This is

similar to the results reported previously for the EFAl-

containing faujasites prepared by conventional incipient

35

wetness impregnation and ion-exchange techniques. Fur-

thermore, these materials do not show a shift of the frequency

of hydroxyl group to lower wavenumbers that is usually

6

1,62

observed in the FTIR spectra of stabilized Y zeolites.

A

fraction of the protons in AHFSY-TMA-Red shows acidity

similar in strength to those in the parent AHFSY zeolite. The

density of BAS in this material is only 10% of that in AHFSY.

For the oxidized materials, this amounts to about 20% of the

original BAS density. The BAS in the oxidized zeolites are

somehow diﬀerent from those in the reduced TMA-modiﬁed

zeolite as their rate of H/D exchange is signiﬁcantly lower. CO

IR spectroscopy conﬁrms the lower density of BAS in the

TMA-modiﬁed zeolites. By using the shift of the OH groups

upon CO perturbation as a criterion for acid strength, no

signiﬁcant diﬀerences in intrinsic proton acidities were

observed. What stands out in the spectra of the reduced

AHFSY-Red zeolite is the presence of strong Lewis acid sites as

Figure 7. Propane cracking activity (TOF(CRC) normalized by the

value for AHFSY) of diﬀerent EFAl-containing faujasite catalysts as a

function of the amount of cationic Al_E_Fper BAS. The activities of

TMA-modiﬁed AHFSY catalysts are shown with red circles. The

concentration of cationic Al_E_Fwas calculated for these samples as

FTIR

c(cationic Al ) = c(Al )

− c(BAS) . The correlation line and

EF

F NMR

the activities of reference USY-8 (blue hexagon) and of AlEF-AHFSY

catalysts prepared by incipient wetness impregnation and ion exchange

of aqueous solution of Al(NO ) (open squares) are adopted from ref

3

XRD

3

5 without modiﬁcation (in this case c(cationic Al ) = c(Al )

−

EF

F

FTIR

c(BAS) ).

present study. Clearly, the oxidized TMA-modiﬁed zeolites

follow the correlation. This is expected because the extraframe-

work Al species are likely present in their oxidic form. When

these species are not completely oxidized as in the reduced

sample, they are much more active than expected based on this

correlation. Acidity characterization showed that the BAS in

−1

evidenced by the CO absorption band at 2222 cm . This band

is attributed to the coordinatively unsaturated Al ions. We

speculate that these ions perturb neighboring BAS. Upon

oxidation, these strong Lewis acid sites completely disappear.

The CO IR spectra also show that all TMA-modiﬁed zeolites

contain weak Lewis acid sites due to agglomerated forms of

1

AHFSY-TMA-Red are strongly polarized ( H NMR) by strong

7

4

“Al−O” species.

Lewis acid Al sites (CO IR). We propose that the unique

1

515

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

enhancement of Brønsted acid catalytic activity of the reduced

TMA-modiﬁed AHFSY zeolite is due to the presence of

strongly polarizing Lewis acid sites resulting from the

REFERENCES

■

(

1) Corma, A. Chem. Rev. 1995, 95, 559.

2) Busca, G. Chem. Rev. 2007, 107, 5366.

(3) Rabo, J. A.; Gajda, G. J. Catal. Rev. Sci. Eng. 1990, 31, 385.

(4) van Santen, R. A.; Kramer, G. J. Chem. Rev. 1995, 95, 637.

(5) Corma, A. Curr. Opin, Solid State Mater. Sci. 1997, 2, 63.

(6) Pine, L. A.; Maher, P. J.; Wachter, W. K. J. Catal. 1984, 85, 466.

7) Sohn, J. R.; Decanio, S. J.; Fritz, P. O.; Lunsford, J. H. J. Phys.

Chem. 1986, 90, 4847.

8) Fritz, P. O.; Lunsford, J. H. J. Catal. 1989, 118, 85.

9) Makarova, M. A.; Dwyer, J. J. Phys. Chem. 1993, 97, 6337.

10) Lago, R. M.; Haag, W. O.; Mikovsky, R. J.; Olson, D. H.;

hydrogenolysis of grafted TMA in H . Because of the strong

2

heterogeneity and structural complexity of the zeolites after

modiﬁcation with TMA, it is not possible to ﬁrmly conclude on

structural or mechanistic aspects of this acidity enhancement.

However, the ﬁnding that CO IR and C D H/D exchange did

(

6

not reveal any special properties of the BAS might suggest that

the Lewis acid sites are involved in the protolytic cracking

mechanism.

(

Hellring, S. D.; Schmitt, K. D.; Kerr, G. T. Stud. Surf. Sci. Catal. 1986,

28, 677.

5

. CONCLUSIONS

(

11) Makarova, M. A.; Bates, S. P.; Dwyer, J. J. Am. Chem. Soc. 1995,

17, 11309.

12) Narbeshuber, T. F.; Brait, A.; Seshan, K.; Lercher, J. A. Appl.

Catal., A 1996, 146, 119.

13) Beyerlein, R. A.; McVicker, G. B.; Yacullo, L. N.; Ziemiak, J. J. J.

Chemical vapor deposition of TMA was explored as an

approach for the introduction of extraframework aluminum in

faujasite-type zeolites to prepare model EFAl-containing zeolite

catalysts. The subsequent decomposition of the grafted

organoaluminum species was investigated in hydrogen and

oxygen atmosphere. Reaction of an extraframework Al free

high-silica AHFSY zeolite with TMA leads to nearly complete

substitution of the bridging hydroxyl groups with Al species.

The reaction, however, does not produce uniform homoge-

neously distributed species. While part of the BAS is replaced,

the highly reactive TMA compound also reacts with the basic

faujasite lattice leading to substantial framework distortions and

even removal of Al from the framework. This is due to the

opening of the Si−O−Al bridges and the formation of stable Si-

CH species. Decomposition of grafted TMA in H leads to a

1

(

Phys. Chem. 1988, 92, 1967.

(14) Sohn, J. R.; DeCanio, S. J.; Lunsford, J. H.; O’Donnell, D. J.

Zeolites. 1986, 6, 225.

(

15) DeCanio, S. J.; Sohn, J. R.; Fritz, P. O.; Lunsford, J. H. J. Catal.

986, 101, 132.

16) Mota, C. J. A.; Bhering, D. L.; Rosenbach, N. Angew. Chem., Int.

Ed. 2004, 116, 3112.

17) Gounder, R.; Jones, A. J.; Carr, R. T.; Iglesia, E. J. Catal. 2012,

86, 214.

18) Agostini, G.; Lamberti, C.; Palin, L.; Milanesio, M.; Danilina, N.;

1

(

2

(

Xu, B.; Janousch, M.; van Bokhoven, J. A. J. Am. Chem. Soc. 2010, 132,

667.

3

2

very active zeolite catalyst for propane conversion. Its activity is

signiﬁcantly higher for propane cracking and dehydrogenation

than that of the parent AHFSY zeolite and a commercial steam-

calcined USY cracking zeolite. Although these eﬀects are less

pronounced when the precursor is oxidized, the propane

conversion rates per proton are still higher than for the

reference samples. Together with the extensive characterization,

we infer that Lewis acidity can signiﬁcantly alter the intrinsic

activity of protons in zeolites. First, strong Lewis acid sites are

directly involved in propane dehydrogenation and, second,

their proximity to BAS results in a pronounced enhancement of

the rates of propane cracking. It cannot be excluded that these

two eﬀects are interrelated by a common mode of propane

activation on Lewis acid sites. The eﬀect is the strongest for the

strong Lewis acid Al sites formed by the reductive treatment of

the TMA-modiﬁed zeolite.

(19) Huang, J.; Jiang, Y.; Marthala, V. R. R.; Thomas, B.; Romanova,

E.; Hunger, M. J. Phys. Chem. C 2008, 112, 3811.

(

20) Jiao, J.; Altwasser, S.; Wang, W.; Weitkamp, J.; Hunger, M. J.

Phys. Chem. B 2004, 108, 14305.

21) Jiao, J.; Kanellopoulos, Wang, W.; Ray, S. S.; Foerster, H.;

Freude, D.; Hunger, M. Phys. Chem. Chem. Phys. 2005, 7, 3221.

22) Li, S. H.; Zheng, A. M.; Su, Y. C.; Zhang, H. L.; Chen, L.; Yang,

J.; Ye, C. H.; Deng, F. J. Am. Chem. Soc. 2007, 129, 11161.

23) Li, S. H.; Huang, S. J.; Shen, W. L.; Zhang, H. L.; Fang, H. J.;

(

Zheng, A. M.; Liu, S. B.; Deng, F. J. Phys. Chem. C 2008, 112, 14486.

(24) Li, S. H.; Zheng, A. M.; Su, Y. C.; Fang, H. J.; Shen, W. L.; Yu,

Z. W.; Chen, L.; Deng, F. Phys. Chem. Chem. Phys. 2010, 12, 3895.

(25) Yu, Z. W.; Zheng, A. M.; Wang, Q.; Chen, L.; Xu, J.; Amoureux,

J. P.; Deng, F. Angew. Chem., Int. Ed. 2010, 49, 8657.

(

26) Babitz, S. M.; Williams, B. A.; Miller, J. T.; Snurr, R. Q.; Haag,

W. O.; Kung, H. H. Appl. Catal., A 1999, 179, 71.

27) Kotrel, S.; Rosynek, M. P.; Lunsford, J. H. J. Phys. Chem. B 1999,

03, 818.

28) van Bokhoven, J. A.; Williams, B. A.; Ji, W.; Koningsberger, D.

C.; Kung, H. H.; Narbeshuber, T. F.; Vinek, H.; Lercher, J. A. J. Catal.

995, 157, 388.

(

1

(

ASSOCIATED CONTENT

■

1

*

S

Supporting Information

(29) Haag, W. O. Stud. Surf. Sci. Catal. 1994, 84, 1375.

2

9

Supplementary XRD, Ar physisorption, FTIR, Si MAS NMR,

(

30) Wei, J. Chem. Eng. Sci. 1996, 51, 2995.

31) Xu, B.; Sievers, C.; Hong, S. B.; Prins, R.; van Bokhoven, J. A. J.

1

27

H-{ Al} TRAPDOR, Al MAS NMR results. This material is

available free of charge via the Internet at http://pubs.acs.org.

Catal. 2006, 244, 163.

32) Ramachandran, C. E.; Williams, B. A.; van Bokhoven, J. A.;

Miller, J. T. J. Catal. 2005, 233, 100.

33) Maesen, T. L. M.; Beerdsen, E.; Calero, S.; Dubbeldam, D.;

Smit, B. J. Catal. 2006, 237, 278.

34) Xu, B.; Sievers, C.; Prins, R.; van Bokhoven, J. A. Appl. Catal., A

007, 333, 245.

35) Almutairi, S. M. T.; Mezari, B.; Filonenko, G. A.; Magusin, P. C.

(

AUTHOR INFORMATION

■

(

Corresponding Author

(

2

(

*

E-mail: e.a.pidko@tue.nl (E.A.P.), e.j.m.hensen@tue.nl

E.J.M.H.).

(

Notes

M. M.; Rigutto, M. S.; Pidko, E. A.; Hensen, E. J. M. ChemCatChem

2013, 5, 452.

The authors declare no competing ﬁnancial interest.

(

36) Coper

Angew. Chem., Int. Ed. 2003, 42, 156.

(37) García-Sanchez, M.; Magusin, P.; Hensen, E. J. M.; Thune, P.

́

et, C.; Chabanas, R.; Saint-Arroman, R. P.; Basset, J. M.

ACKNOWLEDGMENTS

■

́

̈

Financial support was provided by SABIC.

C.; Rozanska, X.; van Santen, R. A. J. Catal. 2003, 219, 352.

1

516

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

ACS Catalysis

Research Article

(

38) Hensen, E. J. M.; Pidko, E. A.; Rane, N.; van Santen, R. A.

Angew. Chem., Int. Ed. 2007, 46, 7273.

39) Almutairi, S. M. T.; Mezari, B.; Magusin, P. C. M. M.; Pidko, E.

A.; Hensen, E. J. M. ACS Catal. 2012, 2, 71.

40) Puurunen, R. L.; Lindblad, M.; Root, A.; Krause, A. O. I. Phys.

Chem. Chem. Phys. 2001, 3, 1093.

41) Van Looveren, L. K.; Geysen, D. F.; Vercruysse, K. A.; Wouters,

B. H.; Grobet, P. J.; Jacobs, P. A. Angew. Chem., Int. Ed. 1998, 37, 517.

42) Bartram, M. E.; Michalske, T. A.; Rogers, J. W.; Mayer, T. M.

Chem. Mater. 1991, 3, 953.

43) Boleslawski, M.; Serwatowski, J. J. Organomet. Chem. 1983, 255,

69.

(74) Zecchina, A.; Platero, E. E.; Arean, C. O. J. Catal. 1987, 107,

244.

(

2

(

44) Pasynkiewicz, S. Polyhedron. 1990, 9, 429.

(

45) Siedle, A. R.; Newmark, R. A.; Lamanna, W. M.; Schroepfer, J.

N. Polyhedron. 1990, 9, 301.

46) Chien, J. C. W.; He, D. J. Polym. Sci., Part A: Polym. Chem. 1991,

9, 1603.

47) Korneev, N. N.; Khrapova, I. M.; Polonskii, A. V.; Ivanova, N. I.;

Kisin, A. V.; Kolesov, V. S. Russ. Chem. Bull. 1993, 1453.

(

2

(

48) Barron, A. R. Macromol. Symp. 1995, 97, 15.

(

49) Harlan, C. J.; Bott, S. G.; Barron, A. R. J. Am. Chem. Soc. 1995,

1

(

17, 6465.

50) Katayama, H.; Shiraishi, H.; Hino, T.; Ogane, T.; Imai, A.

Macromol. Symp. 1995, 97, 109.

(

51) Lee, D. H.; Shin, S. Y. Macromol. Symp. 1995, 97, 195.

52) Sinn, H. J. Macromol. Symp. 1995, 97, 27.

53) Winter, H.; Schnuchel, W.; Sinn, H. Macromol. Symp. 1995, 97,

1

(

3

(

3

(

19.

54) Reddy, S. S.; Radhakrishnan, K.; Sivaram, S. Polym. Bull. 1996,

6, 165.

55) Fusco, R.; Longo, L.; Masi, F.; Garbassi, F. Macromolecules 1997,

0, 7673.

56) U.S. Patent 20110137093 A1, 2011.

57) Sree, S. P.; Dendooven, J.; Koran

́

yi, T. I.; Vanbutsele, G.;

Houthoofd, K.; Deduytsche, D.; Detavernier, C.; Martens, J. A. Catal.

Sci. Technol. 2011, 1, 218.

(

58) Detavernier, C.; Dendooven, J.; Sree, S. P.; Ludwig, K. F.;

Martens, J. A. Chem. Soc. Rev. 2011, 40, 5242.

59) Cairon, O.; Chevreau, T.; Lavalley, J. C. J. Chem. Soc., Faraday

Trans. 1998, 94, 3039.

60) Daniell, W.; Topsøe, N. Y.; Kno

233.

61) Hensen, E. J. M.; Poduval, D. G.; Ligthart, D. A. J. M.; van Veen,

J. A. R.; Rigutto, M. S. J. Phys. Chem. C 2010, 114, 8363.

62) Poduval, D. G.; van Veen, J. A. R.; Rigutto, M. S.; Hensen, E. J.

M. Chem. Commun. 2010, 46, 3466.

63) Li, J.; DiVerdi, J. A.; Maciel, G. E. J. Am. Chem. Soc. 2006, 128,

7093.

64) Hensen, E. J. M.; Pidko, E. A.; Rane, N.; van Santen, R. A. Stud.

Surf. Sci. Catal. 2007, 170, 1182.

65) Van Eck, E. R. H.; Janssen, R.; Maas, W. E. J. R.; Veeman, W. S.

Chem. Phys. Lett. 1990, 174, 428.

(

̈

zinger, H. Langmuir 2001, 17,

6

(

1

(

66) Grey, C. P.; Vega, A. J. J. Am. Chem. Soc. 1995, 117, 8232.

67) Engelhardt, G.; Michel, D. High Resolution Solid State NMR of

Silicates and Zeolites; Wiley & Sons: New York, 1987.

́

68) Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calve, S.;

(

Alonso, B.; Durand, J.; Bujoli, B.; Gan, Z.; Hoatson, G. Magn. Reson.

Chem. 2002, 40, 70.

(

69) Klinowski, J.; Ramdas, S.; Thomas, J. M.; Fyfe, C. A.; Hartman,

J. S. J. Chem. Soc., Faraday Trans. II 1998, 78, 1025.

70) Kentgens, A. P. M.; Iuga, D.; Kalwei, M.; Koller, H. J. Am. Chem.

Soc. 2001, 123, 2925.

71) Czjzek, G.; Fink, J.; Go

(

̈

tz, F.; Schmidt, H.; Coey, J. M.;

Rebouillat, J. P.; Lien

(

́

ard, A. Phys. Rev. B 1981, 23, 2513.

72) Kerber, R. N.; Kermagoret, A.; Callens, E.; Florian, P.; Massiot,

D.; Lesage, A.; Coper

́

et, C.; Delbecq, F.; Rozanska, X.; Sautet, P. J. Am.

Chem. Soc. 2012, 134, 6767.

73) Fu, Q. J.; Ramirez-Cuesta, A. J.; Tsang, S. C. J. Phys. Chem. B

006, 110, 711.

(

2

1

517

dx.doi.org/10.1021/cs400181p | ACS Catal. 2013, 3, 1504−1517

Article Doi

Chemical vapor deposition of trimethylaluminum on dealuminated faujasite zeolite

DOI: 10.1021/cs400181p

Source and publish data:

Authors:

Article abstract of DOI:10.1021/cs400181p

Full text of DOI:10.1021/cs400181p

Products guided by the article

R&D Labs maybe for 34557-54-5

Relevant to this article

Hot Product