Controlling size and yield of zeolite Y nanocrystals using
tetramethylammonium bromide
Brett A.Holmberg,Huanting Wang,Joseph M.Norbeck,Yushan Yan
*College of Engineering––Center for Environmental Research and Technology (CE-CERT)and Department of Chemical and Environmental Engineering,University of California,Riverside,CA 92521,USA
Received 7October 2002;received in revised form 12December 2002;accepted 7January 2003Abstract
In this study,we developed an alternative synthesis technique for the production of small colloidal zeolite Y nanocrystals.By adding tetramethylammonium bromide as a second source of organic template,we were able to de-couple and investigate the e?ect of two important synthesis parameters,tetramethylammonium concentration and anion concentration.Optimizing these two parameters allowed us to hydrothermally synthesize highly crystalline zeolite Y in a disperse nanocrystal form with a controllable particle size of 32–120nm.Crystals hydrothermally synthesized with a 1.00Al 2O 3–4.35SiO 2–2.40(TMA)2O(2OH à)–1.2(TMA)2O(2Br à)–0.048Na 2O–249.00H 2O (T3.6)solution com-position were 45%smaller by volume after 54h of crystallization at 100°C and were obtained with %73%more yield (g zeolitic Al 2O 3+SiO 2/g synthesis Al 2O 3+SiO 2)than crystals synthesized with a 1.00Al 2O 3–4.35SiO 2–2.40(TMA)2-O(2OH à)–0.048Na 2O–249.00H 2O (T2.4)composition.After 216h,T3.6syntheses contained 37nm crystals with 180%higher yield than that of T2.4syntheses at 54h (38nm crystals,0.044g zeolitic Al 2O 3+SiO 2/g synthesis Al 2O 3+SiO 2).The mechanisms involved in the observed crystal size decrease and yield improvement are discussed.A detailed characterization of our zeolite Y nanocrystals is given,including powder X-ray di?raction (XRD),Fourier transform infrared spectroscopy (FT-IR),N 2adsorption,dynamic light scattering (DLS),transmission electron microscopy (TEM),thermogravimetric (TG)analyses,29Si magic angle spinning nuclear magnetic resonance (29Si MAS NMR)and inductively coupled plasma optical emission spectrophotometry (ICP OES)elemental analyses.
ó2003Elsevier Science Inc.All rights reserved.
Keywords:Zeolite;Faujasite;Nanocrystal;FAU;Crystal size
1.Introduction
Zeolite Y is a highly versatile molecular sieve from the faujasite family of zeolites whose 7.4 A
,three-dimensional pore structure and solid acidity
make it useful as a catalyst,ion exchanger,ad-sorbent,and others [1–3].Recently,zeolite nano-crystals were used to construct hierarchical porous materials for many applications such as catalysts,membranes and low-k thin ?lms by our group [4–8],and others [9–11]due to the fact that these materials have high surface areas,fast di?usion characteristics and adjustable porosity [9–16].If simple nanocrystal self-assembly techniques could be applied to large pore acidic zeolites such as *Corresponding author.Tel.:+1-909-787-2068;fax:+1-909-
787-5696.
E-mail address:yushan.yan@92579605e87101f69e319511 (Y.Yan).1387-1811/03/$-see front matter ó2003Elsevier Science Inc.All rights reserved.
doi:10.1016/S1387-1811(03)00271-3Microporous and Mesoporous Materials 59(2003)
13–28
92579605e87101f69e319511/locate/micromeso
zeolite Y,then it is possible to improve existing catalysts that already have widespread applica-tions[1–3].The porosity of hierarchical materials is primarily controlled by varying the precursor nanocrystal size distribution;therefore good con-trol of precursor crystal size is needed.Zeolite Y nanocrystals could also be potentially incorpo-rated into polymer matrices for membrane sepa-ration applications like gas separations[17],and in polymer-exchange-membrane fuel cells[18–21]. For the purpose of developing zeolite-polymer nanocomposite membranes,it is also desirable that zeolite nanocrystals are size adjustable to match polymer-networkdimensions.
Ideally,a high yield nanocrystal synthesis should be used when making nanocrystals for al-most any application.However,high nanocrystal yields are usually favored with longer crystalliza-tion times,which tend to result in larger crystal sizes[22,23].This tendency leads to di?culty in synthesizing crystals for applications that need small crystal size.
Many studies on the synthesis and applications of zeolite Y materials have been carried out[22–25];however the complete understanding of the complex nucleation and crystallization mecha-nisms of zeolite Y growth still eludes researchers. Crystal size control and yield maximization should be possible with a more complete understanding of the mechanisms involved in zeolite Y synthesis.A useful approach to gain understanding of zeolite nucleation and crystallization has been to study the synthesis of zeolite nanocrystals[22,23,26].
Zeolite Y single crystals with an average crys-tal size of80nm were synthesized by optimiz-ing synthesis composition and addition of NaCl [27].The nucleation mechanism of zeolite Y from precursor colloids was revealed by electron mi-croscopy[22].A two-stage varying-temperature technique has been developed to study the nucle-ation/crystallization kinetics of nanosized faujasite zeolites from clear solution[23].
It is well known that organic template plays a key role in the synthesis of zeolite Y nanocrystals. There is a lackof a systematic study on the e?ects of organic template concentration and anion concentrations in zeolite Y nanocrystal growth. The workpresented here is a comprehensive study on the e?ects of template concentration and anion concentration on the synthesis of zeolite Y nano-crystals.
In this study,we introduce tetramethylammo-nium bromide(TMABr)as a second template source in addition to the normal TMAOH tem-plate used in previous recipes.By optimizing the organic template and anion concentrations of the reaction mixture[22,23,27–29],we successfully minimize the size of highly crystalline zeolite Y nanocrystals,while simultaneously increasing the yield.The data presented also provide supporting evidence for solution phase nucleation in the presence of TMABr.Detailed characterization of the zeolite Y nanocrystals is also presented, this includes powder X-ray di?raction(XRD), Fourier transform infrared spectroscopy(FT-IR),nitrogen adsorption,dynamic light scattering (DLS),transmission electron microscopy(TEM), thermogravimetric(TG),29Si magic angle spinning nuclear magnetic resonance(29Si MAS NMR)and inductively coupled plasma optical emission spec-trophotometry(ICP OES).Structural formula of the zeolite Y nanocrystals is also given.
2.Experimental
2.1.Synthesis of zeolite Y nanocrystals
The preparation scheme for hydrothermal syn-thesis of colloidal suspensions of zeolite Y nano-crystals is shown in Fig. 1.First a250mL polypropylene bottle containing a magnetic stir bar was washed with double deionized water(DDI H2O)under ultrasonication.Then appropriate amounts of DDI H2O,25wt%aqueous tetrame-thylammonium hydroxide solution,(TMAOH, SACHEM),and aluminum isopropoxide(97wt%, Aldrich)were added,in that order,and stirred vigorously until the solution became clear. TMABr(98wt%,Fisher)was then added to the clear solution and stirred until completely dis-solved.Aqueous30wt%colloidal silica(Ludox HS-30,SiO2/Na2O?90,wt%SiO2?29.90, wt%Na2O?0.34,Aldrich)was then added to complete the solution,and the bottle was sealed tightly.The?nal molar composition of any
14 B.A.Holmberg et al./Microporous and Mesoporous Materials59(2003)13–28
given synthesis solution was 1.00Al 2O 3:4.36SiO 2:Q (TMA)2O(2OH à):R (TMA)2O(2Br à):0.048Na 2O:249.00H 2O.A typical batch weight composition (T3.6,see Table 1)is as follows:76.5g DDI H 2O,52.3g TMAOH,11.3g TMABr,12.5g aluminum isopropoxide,and 26.2g Ludox HS-30.After each batch was sealed,it was aged for 3days at room temperature (RT)with vigorous stirring,and was then heated with stirring in a 100°C oil bath.Periodic 10mL samples were taken during the synthesis by quickly opening,sampling and resealing the bottle.All samples were washed by three repetitions of centrifugation with rela-tive centrifugal force of 48,500g for 2h;then decanting,and redispersion in DDI H 2O with ultrasonication before analyses preparations were performed.Centrifugations were performed in a Beckman J2-HS centrifuge at 20,500rpm using a JA-21rotor.2.2.Characterization
Dilute washed colloidal suspensions of nano-crystals in ethanol were used to prepare samples for TEM (Philips CM 300at 200kV).TEM work was performed at the Central Facility for Ad-vanced Microscopy and Microanalysis.
DLS (Brookhaven model BI-9000AT digital correlator,Lexel Laser 95Argon Ion Laser)measurements were made using dilute aqueous colloidal suspensions of washed synthesis prod-ucts.Sample temperature was controlled at 24°C via ?uid circulation through the jacket.The argon ion laser was operated at 514nm in light control mode with a power output of 92579605e87101f69e319511ser power output and stability were monitored and con-?rmed for 1h before samples were tested.All measurements were made at a scattering angle of 90°with a 100l m aperture.Sample preparation began by rinsing the sample cuvettes several times with DDI H 2O.The cuvettes were then ultrasoni-cated in DDI H 2O for 15min.After rinsing with more DDI H 2O in the clean room,5mL of DDI H 2O was syringe ?ltered into the cuvettes through a 0.2l m pore size Gelman Science Nylon Acrodisc ?lter,then a washed synthesis product suspension was added drop-wise to the ?ltered DDI H 2O in order to make a suspension that scattered
between
Fig.1.Preparation scheme for zeolite Y nanocrystals.
B.A.Holmberg et al./Microporous and Mesoporous Materials 59(2003)13–2815
0.5and1.0?106counts per second.Generally,the weight loading of the suspension was on the order of0.01wt%.This allowed for the assumption that the viscosity of the suspension was equal to that of DDI H2O.The?nal suspension was clear to the naked eye.The cumulants method was used to determine di?usion coe?cients and mass median diameter of the samples[30,31].All data collected met the conditions necessary to achieve a good?t with the cumulants method using a non-linear least-squares?t of the autocorrelation function. Polydispersity index(PI)was always less than0.3, indicating that the particle size distributions were narrow.Baseline values were always?t better than 0.1%in order to avoid large errors in the PI[32]. No signi?cant improvement of the?t was observed when more than the second moment was used;this indicated that our particle size distributions were approximately symmetrical.

