To synthesize ternary A₁Zn_yZr_zO_n catalysts, we employed the conductive carbon black T100 as the hard template, which was heated at 180˚C overnight before being used. In a typical synthesis, 0.00125 mol of Zn(NO₃)₂∙6H₂O (Sigma-Aldrich, >99.8%), 0.01 mol of ZrO(NO₃)₂∙xH₂O (Sigma-Aldrich, >99.8%), and 0.00125 mol of Ni(NO₃)₂∙3H₂O (Sigma-Aldrich, >99.8%) were dissolved in 40 ml de-ionized water with ultrasonic treatment for 30 mins to obtain a clear and homogeneous solution. Then, 6.0 g of preheated MA100 (Mitsubishi, Japan) was added into the above solution with stirring at the speed of 500 RPM for 2.0 hrs. After that, the mixture was transferred to a ceramic crucible and dried at 110˚C for 12 hrs in oven and further calcined at 400˚C for 4.0 hrs and stay at 550˚C for another 20 hrs to remove the carbon black. The obtained white powders were denoted as Ni₁Zn₁Zr₈O_n catalysts. When Ni(NO₃)₂∙3H₂O was replaced by Fe(NO₃)₃∙9H₂O, Co(NO₃)₂∙3H₂O or Cr(NO₃)₃∙9H₂O, Fe₁Zn₁Zr₈O_n, Co₁Zn₁Zr₈O_n or Cr₁Zn₁Zr₈O_n composite oxide was obtained, respectively.

Several physical techniques were employed to characterize the structure of synthesized catalysts. a) X-ray powder diffraction patterns (XRD) for all the samples were collected on X-ray diffractometer. b) Morphology and size of Cr₁Zn₁Zr₈O_n catalyst were probed by field emission scanning electron microscopy (FESEM, Hitachi S4800). c) Nitrogen adsorption-desorption isotherms were recorded on a Micromeritics ASAP 2460 at 77 K, where the specific surface areas (S_BET), pore volume (V_P) and pore diameter (D_P) were calculated by applying the Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) models to the desorption branches.

The vapor-phase ETIB reactions were evaluated in a tubular quartz reactor (i.d. 5.0 mm). In each run, 200 mg catalyst pelletized catalyst sample (size range: 40 - 60 mesh) was placed at the center of the reactor tube between two layers of quartz wool and were pretreated in N₂ (50 ml/min) at 400˚C for 0.5 h. A mixture of ethanol/H₂O (steam/carbon ratio = 5.0) was then introduced into the evaporator (200˚C) by a syringe pump and directly carried into the reaction system by flowing carrier gas of N₂. An online Shimadzu 2014 Gas Chromatography (GC) with three analyzing channels was used for full product analysis. The carbon balance was checked in each run and found to be higher than 95% in every case. The ethanol conversion and product selectivity were calculated on a per carbon basis and defined by two following equations:

Figure 4 presents the product distributions of ETIB reaction over ternary A₁Zn₁Zr₈O_n and binary Zn₁Zr₈O_n catalysts at 500˚C. One can see that the selectivity of isobutene is 37 % and 39% CO₂ is gained on binary Zn₁Zr₈O_n catalysts. AsNi and Co is introduced forming Ni₁Zn₁Zr₈O_n and Co₁Zn₁Zr₈O_n, respectively. There are no notable changes of product distribution with respect to binary Zn₁Zr₈O_n. But only acetone selectivity decreases or increases. When Fe is introduced forming Fe₁Zn₁Zr₈O_n, isobutene selectivity (8%) decreased dramatically which is far below binary Zn₁Zr₈O_n. In this case a lot of CO₂ (62%) is generated, suggesting that the Fe species favors hydrogen production via steam reforming of ethanol (CH₃CH₂OH + H₂O → CO₂ + H₂). A large amount of H₂ obtained in product effluents also supports this argument. In all ternary catalysts, Cr₁Zn₁Zr₈O_n shows the best isobutene selectivity (51%), while the acetone selectivity increases to 12% and CO₂, CH₄ and ethylene selectivity decreases to 30%, 0.7% and 0.9% respectively, suggesting that ethanol dehydration and cracking are suppressed.

Reaction condition: catalyst weight m = 0.20 g; S/C = 5; x_ethanol = 2.73 mol% (N₂ balance); W/F = 0.0975 g∙s∙ml⁻¹, P = 1 atm., T = 500˚C.

As can be seen from Figure 5, the content of ZnO notably affected the product distribution of bioethanol conversion at 500˚C. When y value varies from 0 to 1, ethylene selectivity decreases and isobutene selectivity increases greatly. At the same time, acetone selectivity increases firstly and then decreases. At y = 2, isobutene selectivity decreases and ethylene shows up again. This suggests that the appropriate content of the zinc addition is very important to the good isobutene selectivity.

From above, we can conclude that the functionalized active sites associated with A, Zn and Zr content in A₁Zn_yZr_zO_n catalyst should be well matched to obtain the highest isobutene selectivity in ETIB reaction. Although the first step, ethanol dehydrogenation is not a difficult reaction, there is still the intense competition between ethanol dehydrogenation and dehydration. The introduction of transition metal species such as Cr can significantly promote the dehydrogenation and suppress the parallel competitive dehydration. The next steps (acetaldehyde-to-acetone and acetone-to-isobutene) need a proper balance between acid and basic sites, which have been evidenced by appropriate Zn addition. Because of the presence of single oxide, which is badly cooperated with other active sites, the introduction of Co or Ni cannot gain better isobutene selectivity. Especially, the introduction of Fe breaks the reaction path and leads to more by-products. Since ZnO is amphoteric oxide, and the excess or the insufficient of its addition will lead to other side reactions and cannot obtain good isobutene selectivity.