Abstract
Introduction
    SBA-15, the quintessential acid-catalyzed non-ionic-copolymer-templated mesoporous silica has served as a support for a variety of earth abundant metals in catalysis, including Al, Ti, V, and Fe (see ref in Y. Li et al. / Microporous and Mesoporous Materials). In particular Fe-SBA-15 shows promise in the upgrading of biomass to Cx-Cy liquid chemicals. (ref here) Iron and its carbide are particularly effective catalysts for the Fischer-Tropsch process for hydrocarbon synthesis. (references?) With growing domestic supplies of natural gas, catalysts which utilize this abundant resource could assitst in the bridge between petroleum derived fuels and biomass sourced alternatives. A current challenge in Fischer-Tropsch catalyis is limiting aggregation and particle growth of Fe during the carburization process, which often occurs in situ during industrial scale Fisch-Tropsch processes. (references) We believe one method to limit this process is the sequestration of Fe within the framework of mesoporous silica. We focus on the direct synthesis of these materials, which involve cocontaminant inclusion of Fe precursor while the silica precursor is introduced. 
    Since the need for a carbon-neutral surrogate to fossil fuel is still relevant where the scale of energy consumption negates electrification, methods to increase the number of Fe sites within these catalysts is a worthy pursuit. Unfortunately the acidic nature of SBA-15 synthesis impedes incorporation of high weight percentages of metal. With the highly acidic (pH < 1) aqueous media used for the preperation of SBA-15, Fe exisits nearly exclusiley as solvated Fe2+ Ions. Additonaly, lack of flanking oxygens as in oxo-metal species, prevent nucleophilic attack on silanol species minimizing direct incopreation of Fe within the silica framework. The acid sensitivity of Fe speciation during the preperation of Fe-SBA-15 has prompted researchers to implement multi-step methods as opposed to "direct-synthesis" in which the metal precursor is added directly to the sythesis media in conjuction with silica precursor.  (ref For example,  iron was first incorporated within mesoporous silica under highly acidic conditions by preperation of Fe zeolites which were cocondensed within forming MSN. Early reports of Fe-SBA 15, by "direct" synthesis often focused on the application of the end-material, with minimal emphasis on Fe incorperation itself. (x wang 2005 x2) Since then a variety of techniques to control the pH while iron is present, which would mitigate the high solubility of Fe in acidic media have been investigated. (ref) Eventually understanding of the effect of Fe/Si ratio was explored from pH 1-6, relativley weak conditions. (ying li 2005) By varying Fe/Si from 1-15 within the gel, Fe/Si within the product spanned 0.043 - 7.41 at pH = 1.5. Further, by setting the gel at conditions which yield Fe/Si = 0.072 and varying pH from 1-5.6, Fe/Si in the product ranged from 0-1. The structure of Fe in these materials often exists as isolated framework species at low Fe/Si ratio, with a distribution of extraframe work clusters, iron oxides, and framework species appearing at high Fe/Si. Many of these early efforts focused on the use of iron nitrate nonahydrate, opperate in weak-acid, or employ ammonium fluoride to modulate the pH.  However, Vinu et al expanded our understanding of the role of x to x ration as a method to surmount of negativley charged silica species and posiitive iron hydroxo compounds under highly acidic conditions used during SBA-15 synthesis. Extrapolating from their previous work towards optimization of Al incorperation within SBA-15, illustrates the role pH plays. High acid content promots formation of aqueous Fe2+ while silanol species also posses apositive charge. The positivley charged silanol units effectivley coordinate with negativley charged ethylene oxide regions of the polymer template, however electrostatic repulsion drives away aqueous metal species. By adjusting the pH above the isoeletric point of silica, a negative charge is developed which iteracts favorably with dissolved metals. (A vinu 2004)
       Despite effective Fe incorperation by pH tuning, other strategies like alternative precursors or co-reactants have not been studied. The majority of the aformentioned materials are prepared from iron nitrate nonahydrate, while a few reports make use of ferric chloride, Fenton's reagent, and ferric oxalate. (ying li 2005,van grieken 2009, khieu dinh/Phu Nguyen 2007, Yon-Mei Liu 2008) The lack of diversity in pecursors justifies a comparative investigation of the effects of precursor on incorporation. In addition, iron content still remains relativley low for direct synthetic methods, with maximum Fe/Si ratios after calcinationof 0.1 and the majority of reports ranging from 47 to 13 (         ying li 2005xq wang 2004a vinu 2005Yon-Mei Liu 2008Bachari 2009van grieken 2009h yan 2016). Here in we explore the direct synthesis of Fe SBA-15 using diammonium phosphate ((NH4)2 HPO4, DHP) additive as an alternative to adjustment of pH. The impact of salt additives on metal incorporation within SBA-15 was first investigated by Budhi et al however  at this time the effect of salt  additive on metal incorperation has not yet been extended to iron. Additon of DHP drastically increases Fe content under the native high acidity of SBA-15 synthesis while preserving high dispersion of metal sites necessary for a sinter reisstant catalyst. We explore the structure of these materials through XRD, SAXRD, 57Fe mossbauer spectroscopy, and EDAX mapping. Experimental results are combined with complex equilibrium calcualtiosn to propose a mechanism of the interaction between iron and dhp leading to increased loading. 
  1. Materials and Methods
    1. Synthesis
All materials prepared vary from a main SBA-15 synthetic recipe. Initially, 4 g of pluronic P-123 are transferred to a 250 mL Teflon bottle. Pluronic P-123 is a tacky gel that can be difficult to handle with conventional techniques, leading to a diminished amount of surfactant reaching the reaction vessel. To address this challenge we found that the surfactant could be gently heated above its melting point and then poured into the Teflon bottle with no observed effect on the final textural or morphological properties. The template,  is dissolved withibn 30 mL of filtered water and x mL of 1.6 M HCl are added. The solution is heated to 80 C stirred at 600 RPM for 1 h. Depending on the desired material, one can now introduce tetraorthoethyl silicate alone or in combination with an iron source. To produce the refece material, 9.8 mg of TEOS were added using a syrring pump operating at. For Fe-SBA-15 prepared without DHP, 0.xx g of the chosen iron precuror are added as soon as the TEOS delivery is initiated. In the case of samples prepared with DHP, first the desired amount of DHP and ferrous ammonium sulfate were ground together within a mortar. This mixture was added to the surfactant solutions as soon as TEOS delivery began. Once the silica source and any desired salts are added to the solution, the reaction is allowed to stir 24 h. Following this initial period, the bottle is sealed and brought to 100 C for an additional 24 h. The solution is then filtered and the recoved  MSN is calcined at 550 C for 5 h.
Characterization
Identification of the prinicpal elements composing the prepared materials was accomplished through a combination of Inductively Couple Plasma Atomic Emission Spectroscopy (ICP-AES) and energy dispersive spectrometry (EDS). (ICP-AES was conducted on a Perkin Elmer Optima 3000 to determine total Fe, P, Si, and S content within the prepared catalysts. Samples were dissolved within 48 wt. % HF and aqua regia, followed by dilution in 5 wt.% HCl. Quality control standards were run every 10 samples (HighPurity QC standards). Transmission Electron Microscopy (TEM) micrographs were acquired on a Phillips CM12 transmission electron microscope operating in bright-field (BF) mode equipped with a W cathode operating at 120 kV accelerating voltage. A FEI TALOS F200X TEM/STEM (HRTEM) equipped with Schottky field emission gun operated at 200 keV was operated in BF and dark-field (DF) modes. The instrument was equipped with a Super-x energy dispersive spectrometer (EDS) for acquisition of spectral images. Images were processed using ImageJ software. All samples were prepared by casting dilute suspensions of as-prepared catalyst in methanol upon lacey-carbon coated Cu grids.
Investigation of the particle morphology, porosity, structural phase, metal distribution, and iron speciation was carried out through 57Fe mossbauer, Nitrogen physsiorption, small angle X-Ray scattering, powder x-ray diffraction, x-ray photoelectron spectroscopy, and TEM. X-Ray Diffraction (XRD) patterns for samples were acquired on a PANALYTICAL PW3040 x-ray diffractometer. θ/2θ scans were acquired between 20-50 ° 2θ with a 4 s dwell time. Samples were prepared by packing of powder within a 1/8” thick, 1” by 1” aluminum holder. Physisorption measurments were carried out on a micrometrics tristar xxx using a 6 h degass at 100 C. Surface areas were derived from the BET equation applied to data below P/P0 < 0.4. Pore width distrbutions were derived from absorption isotherms using BJH theory. Finally interplanar spacings, d, for the expected hexagonal pore network were obtained using the following equation,
a0=2d1003 a_{0}=\frac{2d_{100}}{\sqrt{3}} a0=√32d100
where a is the . SAXS scattering patterns were acquired using a PANalytical x-ray generator with a Cu target and an Anton-Paar SAXSess system. The signal was acquired by exposing of a phosphor image plate detector, which was digitized using a Perkin-Elmer Cyclone readout system and Anton-Parr SAXSquant software. The latter was used to perform the desmearing operation (15 iterations) to correct for the slit geometry of the SAXSess system and also to perform some particle size analyses. Data analysis was conducted on SAXSquant1D and the Irena SAXS package for Igor Pro.2 For absolute intensities (cross section per unit volume, cm-1), methods were developed to measure the transmission of the Cu-Kα x-rays and thickness of each sample*. Transmission values for samples were acquired on a Siemens D500 x-ray diffractometer using Cu-Kα x-rays diffracted from a LaB6 standard. The attenuation of the (110) peak from the LaB6 was measured with and without the sample placed in the beam on the diffracted side.  Samples were prepared by taking a finely divided and homogenized powder consisting of MCM-41 type mesoporous silica and placing it on a commercial acetate adhesive tape. After mounting the sample film onto an aluminum plate with a 3 mm x 30 mm slot, excess powder was gently tapped from the holder assembly. Data processing was performed by extracting the corrected absolute intensity data with SAXSQuant1D. A density of 0.34 g cm-3 was assigned to the SBA-15 sample based on commercial data from Sigma Aldrich. SAXS measurements were made on the commercial acetate tape (which exhibited some scattering, albeit much weaker than sample) and used to correct the intensity.
Mechanism calculations
Visual MINTEQ
Results and Discussion
Non-DHP
To establish the current state of Fe-SBA-15 synthesis in highly acidic conditions we prepared materials from 10 different iron precursors, including the ubiquitous Ferric (III) Nitrate 9 H2O. Literature has deemed Fe incorporation into the silica frame work is controlled by the molar gel ratio of nH2O:nHCl, prompting us to use 1:0.032:0.016:0.364:127 TEOS:Fe2O3:P123:HCl:H2O.2 As seen in table 1 Regardless of the starting Fe compound, we found negligible Fe uptake into the final SBA-15. We also believe due to the high surface area of the final MSN, the presence of Fe precursor in solution also did not extensively impact the formation mesopporous silica. We settled on (NH4)2 Fe (SO4)2 6 H2O ferrous ammonium sulfate due to its compositional similiarty to ammonium heptamolybdate used in our previous investigation of DHP on molybdenum incorperation. Once concern was the impact of additional non-metallic cations or anions on the structure of the final MSN. The effect of ammonium on MSN ha sbeen investigated by . While the impact of anions was considered from two points: first, that the presence of sulfate could impact the state of Fe in solution, and second that the sulfate anion or phosphate anion could interfere with the
S+X-I+, where S+ represents the protonated ethylene oxide groups of P123, I+ is the protonated variant of silica precursos under acidic condistions, and X- represents the coutner anion – usually Cl- from hydrolysis of the acid catalyst, counterion mediated assemblemy mechanism of SBA-15. 3-5 The second concern relates to the previously investigated counterion effect on the lyotophic behavior of MSN formation and we have found no literature reporting on the effect of phosphate anions specifically on MSN preperation, although SBA-15 has been prepared using phosphoric acid as opposed to HCl.6, 7 Since the counterion mediated synthesis mechanism for SBA-15 involves a sufficient induciton period during which anions associate with the charged micelle surface, by delaying the addition of sulfate or phosphate species until after the induction period, we avoided any deleterious effects on the structure of the resulting MSN. At the end of the induction period, the region between micelles is still relativley unpopulated with condesed silica species leaving ample oportunity for metal incorperation