Mechanisms of Adsorption of Aormatic Nitrogen Compounds

MECHANISMS OF ADSORPTION OF AORMATIC NITROGEN COMPOUNDS AND AROMATIC COMPOUNDS ON METAL-ORGANIC FRAMEWORKS (MOFs)

by

JUN DAI

Metal-Organic Frameworks (MOFs) constitute a class of novel porous materials which have attracted significant interest due to their application in separation, storage, catalyst and sensing. Large surface area and porous cavity make MOFs excellent absorbents with huge uptake capacity. In this paper, we studied adsorption mechanisms of adsorption of indole and naphthalene on Basolite F300, Basolite A100 and MIL-100 (Fe) by two complementary spectroscopic methods. Fluorescence spectroscopy and near-UV/Visible diffuse reflectance spectroscopy study demonstrate that naphthalene is quantum confined within the mesoporous cavity in F300. On the other hand, indole is weakly electronically bound to Fe (III) CUS in F300 and forms adsorption complex with F300. Direct spectroscopic proof of adsorption complex is provided by near-UV/Visible spectroscopy and fluorescence spectroscopy. Quenching of ligand-based fluorescence of A100 by indole is suggested and we propose adsorption of indole and naphthalene onto A100 via π-π interaction, spectroscopic proof is provided by fluorescence spectroscopy. 

Table of Contents

Title

Abstract

Acknowledgements and Dedication

Table of Contents

1 Introduction

1.1 Metal Organic Frameworks (MOFs)

1.2 Clean Fossil Fuels

1.3 Nitrogen Aromatic Compounds in Fossil fuels

1.4 Aromatic Compounds in Fossil Fuels

1.5 Methods of Denitrogenation

1.5.1 Microbial Denitrogenation

1.5.2 Hydrodenigrogenation (HDN)

1.5.3 Adsorptive Denitrogenation (ADN)

1.6 Activation of Open Metal Sites of Metal-Organic Frameworks

1.7 Adsorption of Aromatic Compounds and Aromatic N-heterocyclic Compounds on mesoporous MOFs with CUS: MIL-100 (Fe) and F300

1.8 Adsorption of Aromatic Compounds and Aromatic N-heterocyclic compounds on microporous MOFs without CUS: MIL-53 and A100

1.9 Research Objective

2 Experimental

2.1 Metal Organic Frameworks

2.2 Solvents, Aromatic compounds and N-containing compounds

2.3 Activation and Hydration of Metal-organic frameworks

2.4 Fluorescence Spectroscopy

2.5 Near UV-Vis Diffuse Reflectance Spectroscopy (near UV-Vis DRS)

2.6 Model Fuels

2.7 Solid Mixture of Aromatic and Aromatic N-hetrocyclic compounds with MOFs

2.8 Stoichiometric adsorption complex of F300 and naphthalene in eicosane matrix

2.9 Kinetic adsorption of liquid indole on Basolite F300 (FeBTC) in liquid phase

2.10 Stoichiometric adsorption complexes of indole/naphthalene with MOFs

2.11 UV radiation

3 Results

3.1 Spectroscopic Studies of adsorption of naphthalene and indole on mesoporous F300 and MIL-100 (Fe) with CUS

3.1.1 Solid Mixtures of MOFs and organic aromatic compounds

3.1.2 Fluorescence spectrum of pure naphthalene

List of Figures

Figure 1: Structure of MIL-100 (Fe) (see as S1 in supplementary)

Figure 2: Structure of MIL-53 (Al) (see as S2 in supplementary)

Figure 3: Common nitrogen compounds existing in fossil fuels

Figure 4: Pathways for the transformation of quinoline

Figure 5: Fluorescence spectrum of (A) 0.08 M liquid solution of naphthalene (NAP) in n-heptane, λexc = 280 nm, bandwidth: Exc. Slit 5 nm/Emi. Slit 5nm. (B) 0.08 M liquid solution of NAP in n-heptane with multi-Gaussian curve fitting

Figure 6: Fluorescence spectrum of (A) pure naphthalene (NAP) λexc = 280 nm, bandwidth: Exc. Slit 5 nm/Emi. Slit 5nm (B) pure NAP with multi-Gaussian curve fitting

List of Tables

Table 1: MOFs used in this research and structure information

Table 2: Physical property of common nitrogen aromatic compounds existing in fossil fuels

1 Introduction

1.1 Metal Organic Frameworks (MOFs)

Porous materials are well studied and widely used in science and engineering industry. Recently, a novel kind of porous material ‘Metal-Organic Frameworks (MOFs)’ has emerged as potential competitor. MOFs are crystalline hybrid inorganic-organic porous solid formed by chemical metal-linker bounding¹. Metal-Organic Frameworks (MOFs) constitute a class of novel porous materials which have attracted significant interest due to their application in separation, storage, catalyst and sensing. The use of MOF as stationary phase in high resolution GC separation of aromatic hydrocarbons has also gained significant attention². In comparison with adsorption on the MOFs in gas phase which has been well studied and understood, adsorption on the MOFs in liquid phase is much less known. MOFs have extremely large surface area and porous cavity, which make them excellent adsorbents with huge uptake capacity. As a class of coordination compound repeated by metal or metal clusters as SBUs (second building units) and organic linkers, the cavity ranged between that of Zeolite and mesoporous silica³. Regarding the structures, variety of MOFs can display either 2 dimensional or 3 dimensional structures depend on different organic linkers. It is also indicated by J. R. Karra and K. S. Walton that the relation between pore size and guest molecule size plays an essential role in the adsorption⁴. The MOFs being investigated in this research are mesoporous and microporous MOFs. For mesoporous MOFs the pore size can vary from 2 nm to 50 nm, while for microporous MOFs the pore size is less than 2nm. Pore size can be a key in adsorption on MOFs due to micropore filling mechanism², which suggests adsorption capacity depends on molecular cross-sectional area rather than minimum diameter. MOFs are structurally stable under ambient environment. It is reported by many authors that MOFs remain stable under increased temperature and under oxygen. In this research, our investigation mostly focuses on several common MOFS, and their commercial names are Basolite C300, Basolite F300, Basolite A100 and MIL-100 (Fe) (MIL=Materials of Institute Lavoisier). All Basolite C300, Basolite F300 and Basolite A100 are bought commercially, and MIL-100 (Fe) is made with the collaboration of Dr. Jing Li from Rutgers – New Brunswick. Table 1 shows the basic physical information of the MOFs used in this research.

Table 1: MOFs used in this research and structure information

1. MIL-100 family

MIL-100(M) (M=Cr, Fe, Al) refers to a family of mesoporous Metal-Organic Frameworks built up from metal clusters and benzene-1, 3, 5-tricarboxylic (BTC) linkers. These MOFs possess large surface area and have recently attracted tremendous attention for applications in adsorption and separation due to the coordinated unsaturated sites (CUS) which can provide chance for Lewis acid-base interaction between guest molecules and metal ions or metal clusters. In Metal-organic frameworks, metal sites (SBUs) are coordinated with organic linkers. Open metal sites (CUS) may be available on these SBUs. CUS are very important in gas storage, separation, sensing, catalysts, and even biological systems⁶. Considerable amount of research on the adsorption of small molecules have been done recently. Evidence has shown open metal sites have essential influence on adsorption property in MOFs. It is reported that open Cu²⁺ metal sites in HKUST-1 contribute significantly to the high acetylene storage capacity⁷. Moreover, MOFs with CUS such as HKUST-1 are used in chromatography separation due to its specific feature to distinguish strongly and weakly electron donating analytes⁸. In additional, MIL-100 (Fe³⁺, Cr³⁺, Al³⁺) has been shown to have strong preference to adsorb N-heterocyclic compounds (Lewis base) in mixtures⁹¹⁰. Hence, open metal sites can be potential Lewis acid sites, the possible interactions between guest molecules and open metal sites will be examined and discussed in this paper.

In our research, MIL-100 (Fe) will be used to investigate the adsorption mechanisms between guest molecules such as indole or naphthalene and sorbent MOFs. Besides MIL-100, there is a commercially made MOF named Basolite F300, which has a similar chemical composition. Both Basolite F300 and MIL-100 (Fe) are built up by Iron and BTC linkers. However, due to the poor crystallinity, the actual structure of F300 is still not known at this time. Difference in iron content and carbon content in F300 and MIL-100 (Fe) is shown by D. Amarajothi and his co-workers⁵. The iron mass content in F300 and MIL-100 (Fe) are respectively 25% and 21%. The carbon mass content in F300 and MIL-100 (Fe) are respectively 32% and 29%. Pore dimension in F300 and MIL-100 (Fe) are respectively 21.7 Å and 25/29 Å. Structure of MIL-100 (Fe) is shown below:

Figure 1: Structure of MIL-100 (Fe) (see as S1 in supplementary)

2. MIL-53 family

There are many types of MOFs that undergo hydrolysis, while there are other MOFs that remain stable within water. Water stable MOFs start to showing potential applications in drug delivery and imaging¹¹. Many MOFs containing Al as metal sites and amino acids as organic linkers are stable in aqueous solution, and it is also possible to synthesize them in aqueous environment¹². MIL-53(M) (M= metal as Al¹³, Cr¹⁴, Fe¹⁵, In¹⁶, Sc¹⁷) is a common family of microporous MOFs and has excellent water stability.

Figure 2: Structure of MIL-53 (Al) (see as S2 in supplementary)

The one dimensional pore structure of MIL-53(Al) is built by chains of corner-sharing octahedra aluminum clusters AlO₄(OH)₂ and anion form of benzene-1,4-dicarboxylic acid (BDC) as organic linkers connecting infinite numbers of metal clusters. MIL-53 (Al) has rhombic channels which have dimension of 7.3 x 7.7 Å. A huge Langmuir surface area of 1600 m²/g makes MIL-53 (Al) an efficient adsorbent. MIL-53 (Al) has extraordinary thermal stability up to 773 K. It is believed that the structure of MIL-53 (Al) undergoes a reversible structural change during process of adsorbing/desorbing water molecules, and is described as ‘breathing’¹³. The so-called ‘breathing’ process interchanges between large-pore (lp) form (Al(OH)[O₂C-C₆H₄-CO₂]) and narrow-pore (np) form (Al(OH)[O₂C-C₆H₄-CO₂]H₂O) ¹³. The lp form has a dimension of 8.5 x 8.5 Å and is obtained when activated upon high temperature and under high vacuum, which takes away excess free BDC acids, oxygen and water trapped in the cavity during synthesis. The lp form is capable of adsorbing water molecules in vapor at room temperature, as a result, the MIL-53 (Al) shifts to its’ np form. As shown in the formula, the np form contains one trapped water molecule every unit or cavity. Hydrogen bonds are found to form between carboxylic groups on the linkers and adsorbed water molecules. Due to this ‘breathing’ feature, applications on selective adsorption of various compounds in gas phase have already gained interest and being reported¹⁴. The np form of hydrated MIL-53 (Cr) is able to selectively adsorb CO₂ in presence of CH₄ in gas phase. Upon adsorption of CO₂ but not CH₄, MIL-53 (Cr) returns to its lp form with an increased cavity volume of 1522.5 ų from np volume of 1012.8 ų.

In our research, we used Basolite A100 MOF from the BASF, which is commercially available and is equivalent to MIL-53 (Al). Similarly, A100 is build up with AlO₄(OH)₂ octahedral clusters connecting by BDC linkers, and gives a surface area of 1084 m²/g and pore volume of 0.51 cm3/g as determined by N₂ physical-adsorption at 77 K. Additionally, XRD pattern of A100 is proved to resemble that of MIL-53 (Al)¹⁸. G. Blanco-Brieva and his co-workers have studied adsorptive removal of aromatic sulfur compounds from model liquid fuels on thermally activated A100¹⁹.

1.2 Clean Fossil Fuels

Combustion of sulfur and nitrogen containing compounds in fossil fuels will bring adverse effect to public health, environment and economy. It is widely recognized that nitrogen compounds are normally carcinogenic and mutagenic. Separation of heterocyclic compounds from petro-chemical feedstock has become an urgent application leading to clean liquid fossil fuels. Fossil fuel is a primary source of energy on the earth, the intensive use of fossil fuels has raised environmental concerns. Beginning in 2006, the Environmental Protection Agency (EPA) began an ambitious program aiming to reduce the sulfur content of diesel fuels. As of July 11, 2010, the EPA required that the concentration of sulfur content in diesel fuels not exceed 15 ppm²⁰. For certain ultra-clean gasoline, diesel fuels and jet fuels, sulfur level needs to be lower than 1 ppm²¹. Hydrodesulfurization (HDS) of diesel fuel has become an essential research interest, while the presence of nitrogen compounds in middle-distillate oil inhibits the ultra-deep hydrodesulfurization^22232425. In HDS, sulfur compounds are hydrogenated to hydrocarbons and H₂S over catalyst such as Zeolites⁹. However, the nitrogen aromatic compounds in crude oil are found to compete for the active sites on these catalysts to inhibit a deep HDS^26,27. This gives rise to the necessity for denitrogenation in fossil fuels.

Moreover, the combustion of nitrogen compounds in petroleum leads to the formation of NO_x oxides, which is a group of highly reactive and persistent species and contributes directly to acid rain and greenhouse effect. The life time of NO_x oxides in atmosphere is 120 years before being removed or destroyed through chemical reactions. The impact of 1 pound of NO­_x on warming the atmosphere is over 300 times that of 1 pound of carbon dioxide. EPA first set standards for NO₂ in 1971, setting both a primary standard (to protect health) and a secondary standard (to protect the public welfare) at 0.053 parts per million (53 ppb), averaged annually²⁸. Also the presence of nitrogen aromatic compounds in fossil fuels can lead to poisoning of refining catalyst, which will eventually cause a decrease in yield. Thus, denitrogenation is necessary for deep desulfurization and has drawn significant interest around the world due to the increasingly rigid regulations and fuel specifications in many countries.

1.3 Nitrogen Aromatic Compounds in Fossil fuels

Fossil fuels are naturally-formed fuel that contains significant level of heterocyclic aromatic contaminants. There are some representative sulfur aromatic compounds such as benzothiophene (BT) and dibenzothiophene (DBT), representative nitrogen aromatic compounds such as indole and quinoline. Nitrogen aromatic compounds content in crude oil averages around 0.3%. Although the concentration of nitrogen compounds in crude oil is relatively low, the concentration turns out to be higher throughout the petroleum distillation process. Common nitrogen aromatic compounds in fossil fuels are shown in Figure 1.1²⁹ and physical properties of nitrogen aromatic compounds investigated in this research is shown in Table 2.

Nitrogen aromatic compounds in fossil fuels fall into two classes. One is non-basic nitrogen compounds including indole and pyrrole due to the fact that the extra pair of electrons on N is contributed to the π electron cloud and is not available for interaction with acids. While the other class is basic nitrogen compounds such as pyridine, quinoline and their derivatives, in which the lone pair of electrons on N is available as an electron donor.

Indole is one of the most common nitrogen compounds in fossil fuels, and it has been broadly studied. And we choose indole in our research due to the fact that indole is a weakly basic N-containing compound, which means it has a potential of selective adsorption against aromatic compounds and desorption is possible and reasonably convenient.

Figure 3: Common nitrogen compounds existing in fossil fuels

Table 2: Physical property of common nitrogen aromatic compounds existing in fossil fuels

1.4 Aromatic Compounds in Fossil Fuels

Fossil fuels contain significant amount of hydrocarbons, most of them are aromatic compounds and polycyclic aromatic hydrocarbons (PAHs). In this research we choose naphthalene as representative aromatic compounds, and in order to investigate the adsorption mechanism we use fluorescence spectroscopy and UV-Vis diffuse reflectance spectroscopy to investigate electronic interactions upon adsorption onto MOFs. Naphthalene is the most simple and common aromatic compounds in PAHs. And naphthalene is commonly produced in petroleum refining and is then separated from the petroleum. Purification and separation of these aromatic rings is of interest in chemical industry. Naphthalene has a molecular length of 7.2 Å (largest diameter determined by ChemDraw 3D) and is non-polar.

1.5 Methods of Denitrogenation

1. Microbial Denitrogenation

Microbial process is an alternative pathway for denitrogenation. Microorganisms are known to consume natural organic compounds and convert them into carbons and energy, and they are capable of metabolizing certain molecules including nitrogen compounds from fossil fuels. The degradation of quinoline is well-characterized, and the transformation pathways are elucidated by M. Benedik et al²⁹ (shown in Figure 1. 5). The degradation of isoquinoline is less understood but 1-oxo-1,2-dihydroisoquinoline is suggested as initial oxygenated product³⁰. According to the finding of Claus, G, indole is readily degraded via catechol or transformed directly into tryptophan³⁰, while carbazole is relatively more difficult to be degraded. One possible degradation pathway³¹ beginning with angular dioxygenation has been proposed by Ouchiyama, N., which finally enters TCA cycle after conversion to catechol or tryptophan³². Mechanisms for the degradation of other nitrogen compounds such as pyridine, quinoline, acridine and their derivatives are reviewed in detail by J. Kaiser³³. Recent research in microbial denitrogenation has revealed a promising future for application in selective removal of nitrogen- and sulfur-containing petroleum. However, as a major mechanism for the removal and metabolism of organic compounds from the environment, its’ characterization of the enzymes involving in the pathways is still under research²⁹. And it has not been widely applied to the industry yet.

Figure 4: Pathways for the transformation of quinoline

2. Hydrodenigrogenation (HDN)

Hydrodenitrogenation (HDN) is usually used to remove nitrogen compounds in fossil fuels during the refinery process, and Co-Mo catalyst is normally involved in the process. However, the HDS process is accomplished by reacting with hydrogen at high temperature and high pressure. It is energy-intensive, hazardous and costly. Thus, significant amount of researches have been done on HDN in order to reach a goal of being economic and environmentally friendly.

3. Adsorptive Denitrogenation (ADN)

Another promising way to selectively remove the nitrogen compounds in fossil fuels is adsorption on a porous material, Activated carbon^34353637, Zeolites³⁸, HCL-loaded silica-aluminas³⁹, ion-exchange resins⁴⁰, meso-silicas^41,42, Ti-HMSs⁴³, microporous carbon⁴⁴, activated aluminas⁴⁵, Ni-based adsorbents⁴⁵, and NiMOs⁴⁶ have been used for ADN. While recently the use of MOF adsorbents has gained significant interest in adsorption and separation of aromatic and heterocyclic compounds in liquid phase due to its’ high capacity, high selectivity, economic importance and most importantly energy saving ⁴⁷. Moreover, MOFs can even be recycled and effectively reduce the cost. Adsorptive separation via “adsorptive denitrogenation (AND)” is preferred over industrial catalytic HDN¹⁰. Specifically, selective adsorption of N-heterocyclic compounds⁴⁸ in presence of aromatic and aliphatic hydrocarbons in liquid phase⁴⁹ is of interest, which is investigated in this research.

1.6 Activation of Open Metal Sites of Metal-Organic Frameworks

Available CUS in MOFs is essential for the adsorption of Lewis basic compounds. Thus, to fully utilize the open metal sites in MOFs, activation is needed to evacuate the water molecules which are relatively weakly coordinated on these metals. Because MOFs are strongly adsorptive molecules, humidity can be crucial to the availability of open metal sites in MOFs. Specific treatment is necessary at certain temperature under vacuum in order to activate open metal sites. And protection of the sample against humidity is critical during experiments. Interestingly, it is discovered by K. Schlichte and his colleagues that upon activation the color of HKUST-1 would change from light cyan to dark navy⁵⁰. This is confirmed by E. Borfecchia and his working group using UV-Vis, a red shift at LMCT edge and appearance of a shoulder in d-d band at around 600 nm in UV-Vis spectrum is observed and explained by the removal of water⁵¹.

In addition to the activation of MOFs, partial reduction of MIL-100 (Fe) has been reported by H. Leclerc and his group⁵². Certain metal sites in MOF can be reduced, for example, Fe³⁺ in MIL-100 (Fe) can be reduced to Fe²⁺. According to their analysis of oxidation states via IR, outgassing at 423 K for 12 hour will give rise to a greater proportion of Fe²⁺ sites while minority of Fe³⁺ sites is reduced. However, outgassing at 523 K will result in that most Fe³⁺ sites are reduced. Fe³⁺ sites are more Lewis acidic sites than Fe²⁺, it is easier for Fe³⁺ to form coordination bond with nitrogen aromatics, which are good Lewis base.

1.7 Adsorption of Aromatic Compounds and Aromatic N-heterocyclic Compounds on mesoporous MOFs with CUS: MIL-100 (Fe) and F300

Although adsorption of small organic molecules on MOFs both in liquid phase and in vapor, such as toluene, benzene, xylene isomers², etc., has been well studied in the recent years, to our knowledge the adsorption of fuse-ring aromatic hydrocarbons on MIL-100 in liquid phase has not been reported. M. Maes and his co-workers have reported adsorption of indole from heptane/toluene mixture solution on MIL-100 family⁹. Thus, mechanism of adsorption of aromatic compounds versus aromatic N-heterocyclic compounds on F300 has not been studied. According to many studies, the formation of complexes in liquid adsorption is usually assumed, but there is a lack of direct spectroscopic evidence. Adsorption of small aromatic ring compounds (benzene and p-xylene molecules) on MIL-101 has been published by K. Yang, et al². Pore-size filling mechanism is discovered and discussed in this paper, indicating that pore-size and different substrate relates with molecular selectivity of organic compounds. However, the mechanism is still not understood.

Fluorescence spectroscopy is a straightforward method for studying the adsorption mechanism between guest molecules and MOFs. Yet there are rarely papers studying characterization of adsorption of aromatic compounds in liquid or solid system by fluorescence spectroscopy. It is assumed based on fluorescence spectra that Lewis acid-base interaction promotes the adsorption of pyrene on Al₂O₃ from model fuel using octane as solvent⁵³. While the fluorescence spectra was collected using a model sorbent aluminum chloride in methanol instead of actual Al₂O₃ in model fuel, spectroscopic characterization of adsorption complexes formed by MOFs and aromatic compounds or aromatic N-heterocyclic compounds by the fluorescence spectroscopy is still unknown to our knowledge.

Another useful characterization method is near-UV/visible diffuse reflectance spectroscopy (DRS), which is excellent at detecting the variation or shift in electronic states on adsorption complexes forming by metal sites and guest molecules. The DRS was used to identify the interactions between aromatic amines and MOFs⁵⁴ but it has not been used to explore the interactions between metal sites of MOFs and guest molecules as aromatic compounds or aromatic N-heterocyclic compounds.

Thus, we aim to investigate the adsorption of large aromatic compounds on MOFs using two spectroscopic methods as fluorescence spectroscopy and near-UV/visible diffuse reflectance spectroscopy.

1.8 Adsorption of Aromatic Compounds and Aromatic N-heterocyclic compounds on microporous MOFs without CUS: MIL-53 and A100

M. Maes and his colleagues have studied the adsorption of indole and methyl-substituted indole present in model fuel (heptane/toluene at either 80 vol. %/20 vol. % or 20 vol. %/80 vol. % combination) at initial concentration of 0.15 M on thermally activated MIL-53 (Al)⁹. The amount of absorbed indole intensively dropped in model fuel with toluene, which indicates a possible competitive adsorption. While no data of adsorption capacity for indole and substituted indole on MIL-53 (Al) in n-alkane solvent was reported, not enough experimental evidence or computational evidence can be used to prove the mechanism of competitive adsorption. Similarly, adsorption of indole, pyridine, pyrrole and quinolone on thermally activated MIL-53 (Al) has been studied in another paper. In this paper n-octane was used as solvent, however, no data for adsorption capacity of indole on MIL-53 (Al) has been reported⁵⁵. π-π interactions is suggested to be the major force that introduce the adsorption of indole or quinolone onto activated MIL-53 (Al), but no experimental evidence was given⁵⁵. At the present time, no direct spectroscopic characterization of chemical bonds between aromatic or hetero-aromatic adsorbate and MIL-53 has yet been reported. Furthermore, adsorption of naphthalene on MOFs has not been reported to our knowledge.

To investigate the adsorption mechanisms through fluorescence spectroscopy, it is essential to know the origin of the fluorescence from MOF itselft. It was discussed in a review paper by M. Allendorf, C. Bauer, R. Bhakta et al.⁵⁶, that there are five modes for generating fluorescence in MOFs: linker-based, framework metal ions (charge transfer between linker and metal), adsorb