homogeneous and heterogeneous catalysis for biomass

HOMOGENEOUS AND HETEROGENEOUS CATALYSIS FOR BIOMASS

UPGRADING TO PLATFORM CHEMICALS AND END PRODUCTS

by

Angela M. Norton

A dissertation submitted to the Faculty of the University of Delaware in partial

fulfillment of the requirements for the degree of Doctor of Philosophy in Chemical

Engineering

Summer 2020

© Angela M. Norton

All Rights Reserved

HOMOGENEOUS AND HETEROGENEOUS CATALYSIS FOR BIOMASS

UPGRADING TO PLATFORM CHEMICALS AND END PRODUCTS

by

Angela M. Norton

Approved: __________________________________________________________

Eric M. Furst, Ph.D.

Chair of the Department of Chemical and Biomolecular Engineering

Approved: __________________________________________________________

Levi T. Thompson, Ph.D.

Dean of the College of Engineering

Approved: __________________________________________________________

Douglas J. Doren, Ph.D.

Interim Vice Provost for Graduate and Professional Education and

Dean of the Graduate College

I certify that I have read this dissertation and that in my opinion it meets

the academic and professional standard required by the University as a

dissertation for the degree of Doctor of Philosophy.

Signed: __________________________________________________________

Dionisios G. Vlachos, Ph.D.

Professor in charge of dissertation




Signed: __________________________________________________________

Raul F. Lobo, Ph.D.

Member of dissertation committee




Signed: __________________________________________________________

Bingjun Xu, Ph.D.





Signed: __________________________________________________________

Michael Tsapatsis, Ph.D.





Signed: __________________________________________________________

J. Anibal Boscoboinik, Ph.D.


v

This material is based upon the work supported by the Catalysis Center for

Energy Innovation, an Energy Frontier Research Center funded by the U.S.

Department of Energy, Office of Science, Office of Basic Energy under award number

DE‐SC0001004, and the National Science Foundation, award number 1434456.

I extend my sincerest gratitude to my advisor, Dion. You have been

instrumental in my growth and success as a researcher, and I would like to thank you

for every single opportunity you have given me. I hope to emulate your example and

teachings throughout my career.

I extend my thanks to my mentors, Dr. Anibal Boscoboinik and Prof. Michael

Tsapatsis. Thank you for your mentorship during my stay at Brookhaven National

Laboratory. Your expertise and support have been instrumental to my academic

success. I would also like to thank my good friend and colleague, Dr. Lily Cheng, for

always looking out for me and making me laugh.

Last but not least, I would like to express my deepest appreciation to my loved

ones. To my best friend and fiancé, John, I have cherished every moment we have

shared together, and I cannot thank you enough for all you have done to support me.

Thank you for making my dreams come true. To the Ruano-Salguero Family, thank

you for treating me so well and showing me your unconditional love and support.

And, to my parents, John and Grace Norton, and my brother, Jack, thank you for

everything you do. I could not do it without you.

ACKNOWLEDGMENTS

vi

LIST OF TABLES ......................................................................................................... x

LIST OF FIGURES ....................................................................................................... xi ABSTRACT ................................................................................................................. xx

Chapter

1 INTRODUCTION .............................................................................................. 1

1.1 Homogeneous Metal Salt Solutions for Biomass Upgrading .................... 1 1.2 Bio-Lubricant Base Oils Produced by Biomass Upgrading ...................... 5

1.3 Mild Hydrothermal Treatment of Catalysts Involved in Biomass

Upgrading .................................................................................................. 7

1.3.1 Hydrothermal Treatment to Regenerate Catalysts ........................ 8 1.3.2 Hydrothermal Treatment to Anchor Metals .................................. 9

1.4 Thesis Objective and Overview ............................................................... 10

2 HOMOGENEOUS METAL SALT SOLUTIONS FOR BIOMASS

UPGRADING AND OTHER SELECT ORGANIC REACTIONS ................ 13

2.1 Introduction ............................................................................................. 13 2.2 Metal Salt Hydrolysis .............................................................................. 14

2.3 Measurements and Modeling of Elemental Speciation ........................... 16

2.3.1 Experimental Tools ..................................................................... 16 2.3.2 Computational Tools ................................................................... 20

2.4 Metal Salts in Biomass Upgrading .......................................................... 24

3 DIRECT SPECIATION METHODS TO QUANTIFY CATALYTICALLY

ACTIVE SPECIES OF AlCl3 IN GLUCOSE ISOMERIZATION .................. 36

3.1 Introduction ............................................................................................. 36 3.2 Methods ................................................................................................... 38

3.2.1 Speciation Model ......................................................................... 38

TABLE OF CONTENTS

vii

3.2.2 pH Measurements ........................................................................ 39 3.2.3 27Al Quantitative Nuclear Magnetic Resonance (qNMR)

Spectroscopy ................................................................................ 39 3.2.4 Ultrafiltration ............................................................................... 40 3.2.5 Inductively Coupled Plasma Mass Spectrometry (ICP-MS) ....... 40 3.2.6 Dynamic Light Scattering (DLS) ................................................ 40 3.2.7 Catalytic Measurements .............................................................. 41

3.3 Results and Discussion ............................................................................ 41

3.3.1 Model-Predicted AlCl3 Speciation in Aqueous Media ................ 41 3.3.2 Equilibration of AlCl3-HCl Catalyst Solutions ........................... 43

3.3.3 Direct Measurements of AlCl3 Speciation .................................. 46

3.3.3.1 Hexa-Aqua Aluminum (Al3+) ....................................... 47 3.3.3.2 Solid Aluminum ........................................................... 48 3.3.3.3 Hydrolyzed Aluminum Monomers (AlOH2+ and

Al(OH)21+) .................................................................... 48

3.3.3.4 Glucose Conversion and Catalytically Active Species

in Equilibrated Catalyst Solution .................................. 49

3.4 Conclusions ............................................................................................. 52

4 BRANCHED BIOLUBRICANT PRODUCTION THROUGH ALDOL

CONDENSATION ........................................................................................... 53


4.2.1 Materials ...................................................................................... 56

4.2.2 Catalyst Preparation ..................................................................... 57 4.2.3 Reaction Procedures .................................................................... 57

4.2.3.1 Aldol Condensation ...................................................... 57

4.2.3.2 Hydrodeoxygenation (HDO) ........................................ 58 4.2.3.3 Analysis of Products ..................................................... 58

4.2.4 Lubricant Properties .................................................................... 59


4.3.1 Reaction Conditions for Aldol Condensation .............................. 60 4.3.2 HDO to Produce Branched Alkane Base Oil .............................. 64 4.3.3 Lubricant Properties .................................................................... 66

viii

4.4 Conclusions ............................................................................................. 67

5 REVERSIBLE FORMATION OF SILANOL GROUPS IN TWO-

DIMENSIONAL SILICEOUS NANOMATERIALS UNDER MILD

HYDROTHERMAL CONDITIONS ............................................................... 69


5.2.1 MFI Nanosheets ........................................................................... 72

5.2.2 Bilayer Silicate ............................................................................ 73 5.2.3 Scanning Electron and Atomic Force Microscopies (SEM and

AFM) ........................................................................................... 73

5.2.4 Infrared Reflection-Absorption Spectroscopy (IRRAS) ............. 73 5.2.5 X-ray Photoelectron Spectroscopy (XPS) ................................... 74


5.3.1 IRRAS Characterization of 2-D Siliceous Nanomaterials .......... 75

5.3.2 Progressive Hydrothermal Applications to 2-D Nanomaterials .. 77 5.3.3 Single-Step Hydrothermal Application (573 K, 3 mbar H2O, 1

h) .................................................................................................. 83 5.3.4 Mechanistic Insights from H2

18O ................................................ 84

5.4 Conclusions ............................................................................................. 88

5.5 Acknowledgements ................................................................................. 88

6 CONCLUSIONS AND FUTURE WORK ....................................................... 90

6.1 Thesis Summary ...................................................................................... 90 6.2 Future Work ............................................................................................. 92

6.2.1 Metal Salt Catalyzed Glucose Isomerization .............................. 92 6.2.2 Bio-Lubricant Base Oil Production Through Aldol

Condensation ............................................................................... 93

6.2.3 Hydrothermal Conditions Applied to 2-D Siliceous

Nanomaterials .............................................................................. 94

REFERENCES ............................................................................................................. 96

Appendix

A SUPPLEMENTARY INFORMATION FOR CHAPTER 2 .......................... 108

ix

A.1 SnCl4 Speciation (5 mM) in Water at 413 K ......................................... 108 A.2 CrCl3 Speciation (5 mM) in Water at 413 K ......................................... 109

A.3 SnCl4 Species at Varying Concentrations ............................................. 110 A.4 AlCl3 Species at Varying Concentrations .............................................. 111 A.5 CrCl3 Species at Varying Concentrations .............................................. 112

B SUPPLEMENTARY INFORMATION FOR CHAPTER 3 .......................... 113

B.1 Assessment of 3300 HT pH probe and qNMR ...................................... 113

B.2 Aluminum Speciation as Function of Temperature ............................... 115 B.3 Ex Situ and In Situ pH Measurements .................................................. 116 B.4 Equilibration of AlCl3-HCl Catalyst Solutions ..................................... 117

B.5 27Al qNMR Quantification .................................................................... 118 B.6 Dynamic Light Scattering (DLS) .......................................................... 120 B.7 Measured Aluminum Speciation ........................................................... 121 B.8 Initial Rate Constant at 413 K ............................................................... 122

B.9 Initial Rate Constant at 363 K ............................................................... 123 B.10 Identifying the Active Species at 363 K ................................................ 124

C SUPPLEMENTARY INFORMATION FOR CHAPTER 4 .......................... 125

C.1 Micro-Viscometer Apparatus and Standard Measurements .................. 125 C.2 Proposed C-C Coupling Mechanism in C33 Alkane .............................. 126

C.3 GC Chromatographs for Aldol Condensation and HDO Products ........ 127 C.4 HRMS-LIFDI Chromatographs for Aldol Condensation and HDO

Products ................................................................................................. 128 C.5 NMR Results for Aldol Condensation and HDO Products ................... 129

D SUPPLEMENTARY INFORMATION FOR CHAPTER 5 .......................... 133

D.1 SEM and AFM Images of MFI Nanosheets .......................................... 133

D.2 XPS of MFI Nanosheets and Bilayer Silicate ....................................... 134

D.3 Additional IRRAS of MFI Nanosheets and Bilayer Silicate ................. 136

E A FUNDAMENTAL STUDY ON PHOSPHORUS-CONTAINING

ZEOSILS ........................................................................................................ 140

E.1 Background ............................................................................................ 140 E.2 Interaction Between Phosphorus and the Zeosil Framework ................ 140 E.3 Strength of the Acid Sites in P-zeosils .................................................. 147 E.4 Distribution of P-Sites During Steaming ............................................... 152

F PERMISSIONS FOR REPRINT .................................................................... 154

x

Table 4-1: Properties of alkane base oil (YC33 = 50.5% and YC28

= 10.9%) compared to

Group III and Group IV commercial lubricants. ......................................... 67

Table B-1: Data of 27Al qNMR spectra, obtained at 363 K for 24 h, 5 mM AlCl3. ... 118

Table B-2: Data of 27Al qNMR spectra, obtained at 413 K for 24 h, 5 mM AlCl3. ... 119

Table B-3: Total Aluminum Speciation, 5 mM AlCl3, 413 K, measurements obtained

from experiments. ...................................................................................... 121


from experiments. ...................................................................................... 121

Table C-1: Viscosities measured by the micro-viscometer compared to the reported

viscosities. ............................................................................................... 125

Table E-1: IRRAS features and corresponding modes for H3PO4. ............................ 142

Table E-2: IRRAS features and corresponding modes for P(CH3)3. .......................... 144

Table E-3: IRRAS features and corresponding modes for (C2H5)3PO4. .................... 146

Table E-4: The fitting parameters (NML and K) and R2 value for the hydroxyl sites in

de-aluminated BEA, P-BEA, and P-SPP. .................................................. 151

LIST OF TABLES

xi

Figure 1-1: Routes to obtain platform chemicals, HMF and furfural, from sugars found

in lignocellulosic biomass, such as switchgrass, followed by the route to

obtain long-chain ketones from fatty acids found in coconut oil. ............... 4

Figure 2-1: First step in metal salt hydrolysis for a trivalent cation [M(H2O)6]3+........ 15

Figure 2-2: Distribution of AlCl3 speciation (5 mM AlCl3) at reaction temperature

(413 K), calculated using the OLI speciation model. The hydration sphere

of water has been removed for clarity. Redrawn from Norton et al.89 ...... 22

Figure 2-3: Glucose and xylose as platform chemicals with select products shown. .. 24

Figure 2-4: Cellulose hydrolysis followed by glucose isomerization to fructose and its

subsequent transformation to HMF and lactic acid. .................................. 25

Figure 2-5: Proposed mechanism of glucose to fructose isomerization in water. “M”

represents a metal center [e.g., Cr, Al, and Sn (in Sn-BEA)]. Redrawn

from Choudhary et al.14 ............................................................................. 30

Figure 2-6: (a) Effect of HCl on pH for metal chlorides. The intrinsic (self) acidity of

solution (without HCl addition) is a function of the metal. (b) Proposed

active species of Al(III) and Cr(III) chlorides as a function of HCl

concentration. Data points were experimentally deduced Al(OH)21+ from

Norton et al.89 The volcano curve is clearly seen even though quantitative

agreement with the OLI predictions is not great. (c) Proposed active

species of Sn(IV) as a function of HCl concentration. Salt concentration is

5 mM and temperature is 413 K. ............................................................... 35


represents a metal center [e.g., Cr, Al, and Sn (in Sn BEA)]. Redrawn

from Choudhary et al.14 ............................................................................. 37

Figure 3-2: Al (III) ions generated from the dissolution of AlCl3 in aqueous media. .. 42


(413 K), calculated using the OLI software. ............................................. 43

LIST OF FIGURES

file:///C:/Users/Angela/Desktop/Graduate%20UD/Research/Dissertation/Dissertation%20Drafts/Dissertation-Angela-Norton.docx%23_Toc45491183



























xii

Figure 3-4: pH measured ex situ (closed circles) or in situ (open circles) compared to

pH calculated from OLI software (line) at 363 K. .................................... 44

Figure 3-5: Determination of equilibrium from (a) pH, measured ex situ, and (b)

amount of solid formed with time. Samples contained 5 mM AlCl3 and

were cooled from 413 K to 303 K. ............................................................ 45

Figure 3-6: Methodology to quantify aluminum speciation at specified concentrations

of AlCl3 and HCl, temperature, and heating time. .................................... 46

Figure 3-7: 27Al qNMR spectra for Al3+. Experimental conditions: preheat 413 K for

24 h, 5 mM AlCl3. qNMR measurements obtained upon cooling to 303 K.

................................................................................................................... 47

Figure 3-8: Effects of HCl on glucose isomerization rate in AlCl3 solutions (5 mM)

that have been equilibrated at (a) 413 K and (b) 363 K for 24 h prior to

kinetic study. Reaction conditions: glucose 1 wt %, Al to glucose 9 : 100.

................................................................................................................... 49

Figure 3-9: Glucose conversion as a function of measured Al species’ concentrations,

normalized to the maximum observed species’ concentration (see Table B-

3 for observed species’ concentrations). Catalyst solutions were preheated

at 413 K for 24 h prior to kinetic study. Reaction conditions: glucose 1 wt

%, Al to glucose molar ratio of 9 : 100, 413 K. ........................................ 51

Figure 4-1: Renewable approach to produce biolubricant base oil. Base-catalyzed aldol

condensation (AC) of furfural and 12-tricosanone to form a C33 furan

intermediate along with a small fraction of C28 intermediate, followed by

their HDO over an Ir-ReOx/SiO2 catalyst to produce base oils containing

C33 branched alkane as the major component. .......................................... 56

Figure 4-2: Conversion of 12-tricosanone as a function of solvent. Reaction

conditions: furfural (0.227 g) and 12-tricosanone (0.1 g) (mole ratio = 8 :

1), 80 °C, 24 h. Studies performed with no co-solvent, cyclohexane, and

dioxane contained 1 M NaOH (100 µL). Methanol and water studies

contained 1 M NaOH (5 mL). ................................................................... 61

Figure 4-3: Yields of C28 and C33 furan intermediates from the reaction of 12-

tricosanone (0.10 g) with furfural at 80 °C and 1 M NaOH in methanol for

24 h. ........................................................................................................... 62

Figure 4-4: Yields of furan intermediates from the reaction of 12-tricosanone (0.10 g)

with varying amounts of furfural at 80 °C and with 1 M NaOH in



































xiii

methanol (5 mL) for 8 h. Error bars correspond to mean ± standard error

of the mean (SEM) of three independent reactions. .................................. 64

Figure 5-1: Side views of the MFI nanosheets with (a) 7 nm thickness, which

corresponds to 3.5 unit cells and is the visual representation of the

nanosheets used in this work and (b) a 1.5 unit cell, provided to better see

the framework in (a) and which corresponds to the dashed box in (a). (c)

Side view of the bilayer silicate. Red, yellow, and white atoms correspond

to oxygen, silicon, and hydrogen, respectively. Terminal SiOH groups are

present in the MFI nanosheets, but not in the bilayer silicate. .................. 71

Figure 5-2: IRRAS spectra for the SiO2 phonons in (a) MFI nanosheets supported on

Au(111) and (b) bilayer silicate supported on Ru(0001). The SiOH regions

for the MFI nanosheets and bilayer silicate are provided in (c). Scale bars

represent the scale of the y-axis. ............................................................... 76

Figure 5-3: (a) IRRAS of MFI nanosheets supported on Au(111) and (b) quantification

of the IRRAS peak areas for SiOH, taken relative to the initial amount of

SiOH obtained prior to hydrothermal conditions at 300 K and UHV

conditions. (c) IRRAS of bilayer silicate supported on Ru(0001) and (d)

quantification of the IRRAS peak areas for SiOH. All spectra were taken

in the presence of H2O and after 1 h, with the exception of the pre-/post-

H2O and 473 K, 1x10-3 mbar H2O treatments, which had been exposed to

the condition for 60 s and were not in the presence of H2O. The conditions

of each acquisition in (a) and (c) are provided in the x-axes of (b) and (d),

respectively. ............................................................................................... 82

Figure 5-4: (a) IRRAS of MFI nanosheets supported on Au(111). The effect of H2O on

SiOH formation in MFI nanosheets supported on Au(111). Pre-, post-, 573

K under UHV, and 573 K under 3 mbar H2O for 1 h are specified. (b)

Quantification of the IRRAS peak areas for SiOH, taken relative to the

initial amount of SiOH obtained prior to hydrothermal conditions at 300 K

and UHV conditions. ................................................................................. 84

Figure 5-5: A possible reaction pathway showing the formation of SiOH in the MFI

nanosheets with H218O. ............................................................................. 85

Figure 5-6: The effect of H218O on the crystalline structure of (a) MFI nanosheets

supported on Au(111) and (b) bilayer silicate on Ru(0001). Pre-, post-,

573 K under UHV, and 573 K under 3 mbar H218O are specified. ........... 87

Figure A-1: (a) Distribution of hydrolyzed SnCl4 species (5 mM SnCl4) at 413 K,

calculated using the OLI speciation model. The hydration sphere of water





































xiv

has been removed for clarity. (b) Zoomed in profile of the dominate

SnO2(solid) species. .................................................................................... 108

Figure A-2: Distribution of CrCl3 speciation (5 mM CrCl3) at 413 K, calculated using

the OLI speciation model. The hydration sphere of water has been

removed for clarity. ................................................................................. 109

Figure A-3: Speciation profiles obtained using the OLI speciation model at 413 K for

(a) SnO2(solid), (b) SnO2(solid) zoomed in profile, (c) Sn(OH)4, (d)

Sn(OH)31+, (e) Sn(OH)2

2+, and (f) Sn(OH)51- as a function of HCl

concentration, where SnCl4 concentration is 5, 15, and 30 mM. zoomed in

profile. ..................................................................................................... 110


(a) Al(OH)21+, (b) Al(OH)2+, (c) AlO(OH)solid, and (d) Al(OH)3 as a

function of HCl concentration, where AlCl3 concentration is 5, 15, and 30

mM. ......................................................................................................... 111


(a) CrOH2+, (b) CrO1+, (c) CrCl2+, and (d) Cr(OH)3(solid) as a function of

HCl concentration, where CrCl3 concentration is 5, 15, and 30 mM. ..... 112

Figure B-1: pH measurements obtained on the 3300 HT pH probe (circles) as a

function of temperature for the (a) pH = 1.678 and (b) pH = 3 buffer

solutions. The solid lines represent the expected pH. Solution temperature

control was applied. ................................................................................. 113

Figure B-2: Concentration of Al3+ measured by 27Al qNMR (closed circles) as a

function of HCl concentration, along with OLI model predictions (line).

Conditions: 303 K, 5 mM AlCl3 with HCl concentration between 0 and

200 mM. .................................................................................................. 114

Figure B-3: Effect of temperature on the distribution of (a) Al3+, (b) AlO(OH), (c)

Al(OH)2+, and (d) Al(OH)21+ in 5 mM AlCl3 as a function of HCl

concentration, calculated using the OLI software. .................................. 115

Figure B-4: pH measured ex situ (closed circles) or in situ (open circles) compared to

pH calculated using the OLI speciation software (lines) at (a) 303 K and

(b) 413 K. ................................................................................................ 116

Figure B-5: Average particle diameter (nm), obtained from DLS, as a function of time.

Experimental conditions: 5 mM AlCl3 and HCl (specified), preheated at

(a) 413 K and (b) 363 K for 24 h and cooled to room temperature......... 117



































xv

Figure B-6: (a) pH and (b) amount of solid vs. time at different HCl initial

concentrations. Samples contained 5 mM AlCl3 and were cooled from 363

K to 303 K prior to measurements. ......................................................... 117

Figure B-7: 27Al qNMR spectra for Al3+. Experimental conditions: preheat 363 K for

24 h, 5 mM AlCl3. qNMR measurements obtained at 363 K. ................. 118

Figure B-8: DLS spectra of pre-filtered and post-filtered catalyst solution.

Experimental conditions: 5 mM AlCl3, 0 mM HCl, preheat at 413 K (24 h)

and cooled to room temperature. ............................................................. 120

Figure B-9: (a) –ln(cglucose/cglucose,0) vs. time, (b) – (f) conversion of glucose and carbon

balance vs. time. Catalyst solutions were preheated at 413 K for 24 h prior

to kinetic study. Reaction conditions: glucose 1 wt %, Al to glucose molar

ratio 9 : 100, 413 K, , and either (b) 0 mM, (c) 3 mM, (d) 10 mM, (e) 20

mM, or (f) 44 mM HCl. ........................................................................... 122

Figure B-10: (a) –ln(cglucose/cglucose,0) vs. time, (b) – (f) conversion of glucose and

carbon balance vs. time. Catalyst solutions were preheated at 363 K for 24

h prior to kinetic study. Reaction conditions: glucose 1 wt %, Al to

glucose molar ratio 9 : 100, 363 K, and either (b) 0 mM, (c) 3 mM, (d) 10

mM, (e) 20 mM, or (f) 44 mM HCl. ....................................................... 123

Figure B-11: Glucose conversion as a function of measured Al species’

concentrations, normalized to the maximum observed species’

concentration (see Table B-4 for observed species’ concentrations).

Catalyst solutions were preheated at 363 K for 24 h prior to kinetic study.

Reaction conditions: glucose 1 wt %, Al to glucose molar ratio 9 : 100,

363 K. ...................................................................................................... 124

Figure C-1: (a) Illustration and (b) photo of micro-viscometer (Cannon, calibrated

model #: 9722-H62) used to obtain kinematic viscosities (KVs) in

accordance with ASTM D445. ................................................................ 125

Figure C-2: (a) Carbon-carbon (C-C) cracking in the tertiary carbon positions of the

C33 alkane to produce alkane byproducts. (b) C-C cracking in secondary

carbon positions; the dashed blue line is another possible route to obtain

the C14 alkane instead of the C15 alkane. ................................................. 126

Figure C-3: (a) GC spectra of aldol condensation (AC) product obtained from a

reaction of furfural (0.45 g) and 12-tricosanone (0.10 g). Potential isomers

are highlighted in the panels on the right. (b) GC spectra of product

obtained after HDO of aldol condensation product. AC products are C28

and C33 furan isomers (0.40 g). HDO products are primarily chiral isomers





































xvi

of the C28 and C33 alkanes. Eicosane (0.10 g) is used the internal standard

to quantify products, post-reaction, in (a) and (b). m/z was obtained

separately, using GC-MS. ....................................................................... 127

Figure C-4: HRMS-LIFDI to detect the mass fragments of products after (a) aldol

condensation and (b) hydrodeoxygenation reactions. Solvents for both

reactions were removed prior to analysis, and samples were prepared in

dichloromethane (1 mg/mL). m/z was obtained from HRMS-LIFDI. .... 128

Figure C-5: 1H NMR spectrum to characterize aldol condensation product. Sample

was prepared in CDCl3 (1 mg/mL). The predominant product is the C33

furan, followed by the C28 furan, and their isomers. The highlighted bonds

in C33H50O3 are referencing the hydrogen atoms. ................................... 129

Figure C-6: 13C NMR spectrum to characterize aldol condensation product. Sample

was prepared in CDCl3 (1 mg/mL). The highlighted bonds in C33H50O3 are

referencing the carbon atoms. ................................................................. 130

Figure C-7: 1H NMR spectrum to characterize hydrodeoxygenation product. Sample

was prepared in CDCl3 (1 mg/mL). The highlighted bonds in C33H68 are

referencing the hydrogen atoms. High molecular weight oxygenates are

likely present, as indicated by the chemical shifts at ~3.7 ppm. ............. 131

Figure C-8: 13C NMR spectrum to characterize hydrodeoxygenation product. Sample

was prepared in CDCl3 (1 mg/mL). The highlighted bonds in C33H68 are

referencing the carbon atoms. Chemical shifts above 60 ppm may

correspond to the carbons associated with the unidentified oxygenates. 132

Figure D-1: (a) SEM and (b) AFM images of MFI nanosheets on Au(111) taken in air.

In (a), the MFI nanosheets almost entirely cover the surface (coverage ~1),

with some overlapping regions between corners of some of the nanosheets

(which appear darker gray in the image). The color scale on the right-hand

side of (b) shows the height differences (in Å) on the surface. While most

of each nanosheet surface is flat, with a thickness of ~7 nm, they have a

thicker region in the middle corresponding to the seed material used for

their growth. Note the SEM and AFM images were taken on two separate

regions of the material. ............................................................................ 133

Figure D-2: The XPS of MFI on Au(111) was taken in ultra-high vacuum (UHV) at

300 K, prior to hydrothermal treatment. ................................................. 134

Figure D-3: The XPS of bilayer silicate was taken in UHV at 300 K, prior to

hydrothermal treatment. .......................................................................... 135




































xvii

Figure D-4: (a) IRRAS of MFI nanosheets supported on Au(111). The effect of H2O

on SiOH formation in MFI nanosheets supported on Au(111). Pre-, post-,

573 K under UHV, and 573 K under 3 mbar H2O for 1 h are specified. (b)



and UHV conditions. The peak areas were obtained from the spectra

plotted in absorbance; not much difference was observed between

absorbance and transmission spectra. ...................................................... 136

Figure D-5: The effect of H2O on the crystalline structure of (a) MFI nanosheets

supported on Au(111) and (b) a polymorphous bilayer silicate supported

on Ru(0001). Pre-, post-, and during (573 K, 3 mbar) H2O treatment are

labeled. .................................................................................................... 137

Figure D-6: IRRAS of the effect of D2O at varying temperatures and pressures of D2O

for (a) MFI nanosheets supported on Au(111) and (b) polymorphous

bilayer silicate on Ru(0001). IRRAS is shown for: (i) pre-exposure to

D2O at 300 K, UHV, (ii) 10-3 mbar at 473 K for 60 s, (iii) 10-2 mbar at 473

K for 1 h, (iv) 10-1 mbar at 473 K for 1 h, (v) 1 mbar at 473 K for 1 h, (vi)

3 mbar at 473 K for 1 h, (vii) 3 mbar at 573 K for 1 h, and (viii) post-

exposure to D2O at 300 K, UHV. Spectra in (ii) through (viii) in (a) and

(b) use (i) as a reference. As a result of this, when plotted in % reflectance

units, features pointing up indicate the disappearance of pre-existing

species, while features pointing down show the appearance of new

species. .................................................................................................... 138

Figure D-7: (a) IRRAS of bilayer silicate supported on Ru(0001). Pre-, post-, 573 K

under UHV, and 573 K under 3 mbar H2O for 1 h are specified. (b)



and UHV conditions. ............................................................................... 139

Figure E-1: IRRAS spectra for the SiO2 phonons in the bilayer silicate supported on

Ru(0001). ................................................................................................. 141

Figure E-2: IRRAS spectra for the bilayer silicate supported on Ru(0001) exposed to

H3PO4 at 141 K. The feature at 1300 cm-1 (unlabeled) corresponds to the

SiO2 phonons in the bilayer silicate. ....................................................... 142


(a) H3PO4 at 141 K. (b) H3PO4 fully desorbs at 283 K. ......................... 143




































xviii


P(CH3)3 at 141 K. The feature for the bilayer silicate (1300 cm1) overlaps

with the C-H deformation bands. ............................................................ 144


(C2H5)3PO4 at 141 K. The feature at 1300 cm-1 (unlabeled) corresponds to

the SiO2 phonons in the bilayer silicate. ................................................. 145


(C2H5)3PO4 at 110 K. (C2H5)3PO4 fully desorbs at 210 K. ..................... 147

Figure E-7: P-BEA (Si/P = 27), P-SPP (Si/P = 27), and de-aluminated BEA

(deAlBEA) spectra, acquired at 300 K. Catalysts were pre-heated at 623 K

under vacuum (1×10-5 Torr) for 12 h. Spectra were acquired upon cooling

to 300 K. .................................................................................................. 148

Figure E-8: De-aluminated BEA was pre-heated at 623 K under vacuum (1×10-5 mbar)

for 12 h, followed by exposure to varying levels of D2O at 300 K for 1 h.

The corresponding pressures include: (i) 3.8×10-7 Torr, pre-D2O, (ii)

4.0×10-4 Torr D2O, (iii) 2.3×10-1 Torr D2O, (iv) 2.3 Torr D2O, and (v)

3.9×10-4 Torr, post-D2O. SiOH (3732 cm-1) gradually exchanges to SiOD

(2748 cm-1). Likewise, SiOH nests (3472 cm-1) gradually exchange to

SiOD nests (2604 cm-1). .......................................................................... 149

Figure E-9: P-BEA was pre-heated at 623 K under vacuum (1×10-5 mbar) for 12 h,

followed by exposure to varying levels of D2O at 300 K for 1 h. The

corresponding pressures include: (i) 2.4×10-7 Torr, pre-D2O, (ii) 11.3×10-3

Torr D2O, (iii) 1.9×10-1 Torr D2O, (iv) 4.0 Torr D2O, and (v) 5.6 Torr

D2O, (vi) 6.7 Torr D2O, (vii) 7.5×10-4 Torr, post-D2O. SiOH (3745 cm-1)

gradually exchanges to SiOD (2756 cm-1). POH (3666 cm-1) gradually

exchanges to POD (2692 cm-1). .............................................................. 150

Figure E-10: P-SPP was pre-heated at 623 K under vacuum (1×10-5 mbar) for 12 h,

followed by exposure to varying levels of D2O at 300 K for 1 h. The

corresponding pressures include: (i) 3.4×10-7 Torr, pre-D2O, (ii) 3.8×10-3

Torr D2O, (iii) 3.2×10-2 Torr D2O, (iv), 1.4×10-1 Torr D2O (v) 1.0 Torr

D2O, (vi) 2.0 Torr D2O, (vii) 3.0 Torr D2O, (viii) 4.4 Torr D2O, (ix) 5.0

Torr D2O, (x) 6.3 Torr D2O, (xi) 4.0×10-4 Torr, post-D2O . SiOH (3745

cm-1) gradually exchanges to SiOD (2760 cm-1). POH (3670 cm-1)

gradually exchanges to POD (2706 cm-1). .............................................. 150

Figure E-11: The band height decrease (N, y-axis), obtained from the transmission IR

spectra for the SiOH and POH groups in P-BEA, P-SPP, and de-





































xix

aluminated BEA, were plotted as a function of increasing D2O pressure

(Torr). The data were fit with a Langmuir-type equation. ...................... 151

Figure E-12: (a) Transmission IR spectra collected for P-BEA (Si/P = 3), acquired

during pre-exposure to H2O at 1×10-5 mbar (pink), 10 Torr H2O (blue),

and post-exposure to H2O at 1×10-5 mbar (orange). All spectra were

acquired at 523 K. (b) Zooming in on the hydroxyl region of the IR

spectra for P-BEA, SiOH and POH groups appear at 3740 cm-1 and 3665

cm-1, respectively. Prior to acquisition, the catalyst sample was heated

under vacuum (1×10-5 mbar) at 673 K for 4 h. ....................................... 152


during pre-exposure at 523 K and 10 Torr H2O as a function of time. (b)

Zooming in on the hydroxyl region of P-BEA, SiOH and POH groups

appear at 3740 cm-1 and 3665 cm-1, respectively. Prior to acquisition, the

catalyst sample was heated under vacuum (1×10-5 mbar) at 673 K for 4 h.

................................................................................................................. 153

Figure F-1: Permission for Chapter 1 and Chapter 2 ................................................. 155

Figure F-2: Permission for Chapter 4 ......................................................................... 157


















xx

Today, transportation fuels and lubricants, plastics, fabrics, and many other

products are derived from petroleum, a nonrenewable feedstock and contributing to

greenhouse gas emissions. Bioderived, renewable feedstocks can mitigate the

environmental impact of manufacturing commodity products, while providing

comparable or even superior properties to their petroleum-derived counterparts.

Among the renewable carbon sources, nonedible lignocellulosic biomass is one of the

most promising feedstock alternatives; however, its high oxygen content remains the

most significant barrier for its conversion into low-oxygen containing fuels and

chemicals traditionally derived from petroleum. As a result, multiple processes have

been studied for oxygen removal, such as the preparation of oxygenated furans from

biomass-derived sugars. From there, oxygenated furans can undergo additional

transformations to form high-value products, such as lubricants. All of these

transformations require a suitable homogeneous or heterogeneous catalyst to enable

highly selective and energy efficient processes. To this end, the objective of this thesis

is to investigate selected catalytic systems and processes to obtain valuable products

from biomass.

Chapter 3 of this thesis begins by identifying and investigating the molecular

species derived from the homogeneous metal salt, AlCl3, that enable the conversion of

sugar (glucose, C-6) in biomass to 5-hydroxymethylfurfural (HMF). While

homogeneous metal salts have been shown to catalyze sugar chemistries, direct

experimental evidence in support of a specific catalytic species remains elusive. Here,

ABSTRACT

xxi

direct speciation measurements are coupled with kinetics to provide convincing

evidence that [Al(H2O)4(OH)2]1+ is the active species for glucose conversion in water.

A speciation model is used to predict aluminum species as a function of composition,

while simultaneously an experimental protocol is used to quantify the various

aluminum species. Linear scaling between the glucose conversion rate and the

speciation measurements at sufficiently high temperatures indicates that the

[Al(H2O)4(OH)2]1+ complex is the active species in glucose conversion. Knowledge of

the active species can help improve future catalyst development for this and other

reactions.

In Chapter 4, biomass-derived platform chemicals are used to produce bio-

lubricant base oils. Our strategy involves coupling 12-tricosanone, obtained from

bioderived fatty acids, with furfural, a C-5 species, obtained from hemicellulose, to

form a highly branched bio-lubricant base oil. I show that the viscous properties of the

final product are comparable to commercial petroleum-derived Group III and Group

IV base oils.

Another major challenge in biomass conversion is catalyst deactivation.

Despite its importance, in situ characterization of catalytic materials is often difficult.

In Chapter 6 of this thesis, the effects of mild hydrothermal treatment on siliceous

nanomaterials are monitored in situ by infrared reflection absorption spectroscopy

(IRRAS). Well-ordered, siliceous materials called zeosils are employed as supports for

metal catalysts in biomass upgrading, but require frequent regeneration to remove

carbonaceous by-products from their pores. Monitoring the effects of mild

hydrothermal conditions, in situ, on siliceous materials has been challenging to

observe by IRRAS, which requires flat surfaces and benefits from electrically

xxii

conductive substrates. To address this challenge, I use 2-D siliceous nanomaterials

deposited on metal single crystals. In Chapter 6, it is shown that elevated temperatures

and water pressures increase the formation of silanol (SiOH) groups in the MFI

nanosheets, but do not change the polymorphous bilayer silicate. The effects are fully

reversible in the MFI nanosheets. The implications shown here may provide insights

into the effects of mild hydrothermal treatment applied to siliceous 3-D materials,

which are challenging to study using surface science.

1

INTRODUCTION

The vast majority of consumer products, including automotive fuels and

lubricants, plastics, and synthetic fabrics, are obtained from petroleum, a

nonrenewable feedstock, whose processing and combustion contributes to greenhouse

gas emissions. The use of bioderived, renewable feedstocks to produce these or similar

products can reduce the environmental impact, while providing materials with

comparable or even superior properties to their petroleum-derived counterparts.

Among the renewable energy sources, nonedible lignocellulosic biomass is one of the

most promising petroleum alternatives, as it is abundant, contains carbon, the

building-unit for many products of interest, does not interfere with food consumption,

and can result in a nearly “closed carbon balance” through CO2 capture during

photosynthesis. The U.S. Department of Energy (DOE) projects that the U.S. has the

potential to produce, by 2040, 1 billion dry tons of non-food biomass per year,

resulting in 50 billion gallons of biofuel and 50 billion pounds of bio-based chemicals

and bio-products annually,1 placing the U.S. in an advantageous position in the global

energy and climate landscape. To realize this projection, the high oxygen content in

biomass must be overcome.

1.1 Homogeneous Metal Salt Solutions for Biomass Upgrading

Nonedible lignocellulosic biomass includes switchgrass, cornstover,

agricultural waste, or any form of plant-derived material that does not interfere with

Chapter 1

2

food consumption. The three major components of lignocellulosic biomass are

cellulose, a polymer of glucose (C-6), hemicellulose, a polymer of both glucose (C-6)

and xylose (C-5), and lignin, a highly cross-linked polymer of aromatic compounds.2,3

The high oxygen content in biomass is one of the most significant barriers for its

conversion into low-oxygen containing fuels and chemicals traditionally derived from

petroleum. As a result, multiple processes have been studied for oxygen removal in

lignocellulosic biomass. For example, pyrolysis, a thermal process (>400 °C), breaks

down biomass, but typically results in unselective breaking of chemical bonds and is

highly energy intensive. In contrast, enzymatic catalysis employs low temperatures but

is rather limited as it is slow, has a narrow pH operating window, and requires a rather

purified reaction feedstock, aspects that restrict the cost-effective coupling of upstream

and downstream processes. Promising alternatives entail homogeneous and

heterogeneous acid catalysts to enable multiple reactions in a single-pot at low

temperatures (<170 °C).4–7 While heterogeneous catalysts are often favored over their

homogeneous counterpart, insoluble polymers known as humins,8 forming during

reaction can deposit in the catalyst pores, requiring frequent catalyst regeneration. In

this respect, homogeneous catalysts have clear advantages over thermal, biological,

and heterogeneous ones.

The reaction pathway to convert biomass to chemicals involves the conversion

of the cellulose and hemicellulose fractions into monomeric sugars, glucose (C-6) and

xylose (C-5), respectively. Glucose and xylose then undergo isomerization to fructose

and xylulose, followed by dehydration to form 5-hydroxymethylfurfural (HMF) and

furfural. HMF and furfural are “platform chemicals” for many products, including

biodiesel, biodegradable plastics, and biopharmaceuticals (Figure 1-1).9 Since 1921,

3

the Quaker Oats process has been established to convert the pentose (C-5) fraction of

agricultural waste to furfural,10 but the industrial production of HMF is not yet

economically viable, as HMF is derived from glucose, which often degrades in acidic

solution.9 Fructose can much more easily be dehydrated to HMF, but it is not as

naturally abundant as glucose. Consequently, research efforts have been focused on

improving the conversion of glucose to fructose.

Following the pioneering work of Davis and co-workers, who studied

heterogeneous Lewis acid catalysts, e.g., Sn-BEA, metal salts were exploited as

homogeneous Lewis acid catalysts for isomerization.11 In one of the first studies

involving metal salts, Heeres and co-workers found that Al3+, Cr2+, and Zn2+ resulted

in the highest conversion of glucose to either HMF or lactic acid, a by-product of

glucose conversion.12 Subsequently, Zhang and co-workers discovered that CrCl3 in

ionic liquids performed tandem catalysis of the isomerization and dehydration

chemistries.13 Vlachos and co-workers showed that the mechanism of glucose to

fructose, catalyzed by CrCl3 or AlCl3, was analogous to that of heterogeneous Sn-BEA

zeolite.14 It was proposed that the active species for both homogeneous and

heterogeneous catalysts likely exist as a bifunctional Lewis-acidic/Brønsted-basic

site;14 however, direct experimental evidence in support of the catalytic species has so

far been lacking. The mechanistic similarities observed between homogeneous and

heterogeneous catalysts suggest that knowledge from homogeneous catalysts, which

are easier to study and do not invoke synthesis challenges, could be applied to the

identification and improvement of heterogeneous catalysts.

Designing better catalysts requires understanding the active species responsible

for the chemistry. In a succeeding study by Vlachos and co-workers, speciation

4

modeling and glucose isomerization kinetics were combined to propose that CrCl3

hydrolyzes in water to form a hydroxy species known as [Cr(H2O)5(OH)]2+, which is

active in glucose isomerization.7 In another work, Hu and co-workers performed

tandem electrospray ionization mass spectrometry (ESI-MS/MS) to propose, due to its

abundance, [Al(H2O)4(OH)2]1+ is the active species for glucose isomerization.15 Given

the fact that ESI-MS/MS may alter speciation during the measurement16 and that

observed species are often spectators rather than the active ones, the active aluminum

species still remain(s) unknown.

Figure 1-1: Routes to obtain platform chemicals, HMF and furfural, from sugars found

in lignocellulosic biomass, such as switchgrass, followed by the route to obtain long-

chain ketones from fatty acids found in coconut oil.

FATTY ACIDS

Glucose Fructose

Xylose Xylulose Furfural

HMF

COCONUT OIL

SUGARS

SWITCH GRASS

Lauric acid 12-Tricosanone

R = C10H21

C-6 Sugars

C-5 Sugars

5

1.2 Bio-Lubricant Base Oils Produced by Biomass Upgrading

In this section, the focus shifts from producing platform chemicals to using

them as reagents to make bio-lubricant base oils. Lubricants are widely used in

industrial and aviation machineries, automobiles, agricultural equipment, marine

vessels, refrigeration compressors, and many other applications and represent an over

$60 billion global chemical enterprise. Base oils are key components (typically, 70 –

90 wt. %) of commercial formulated lubricants and account for 75% of lubricant cost.

Base oils are also key components in the formulations of personal care products,

greases, and the like. According to the American Petroleum Institute (API), there are

five categories of mineral base oils (Groups I – V).17 Base oils from Groups I through

III are obtained by solvent-refining, distillation, or hydro-processing of petroleum.18

Base oils from Group IV have undergone chemical upgrading; for example, 1-decene

from petroleum undergoes oligomerization to form poly-α-olefins (PAOs).19 Finally,

Group V includes all other oils.17 ExxonMobil Basestocks 2018 Pulse Report suggests

Group III+ oils will experience the greatest increase in demand over the next 10 years

(+4% increase) owing to their high fuel efficiency and quality.20 Nevertheless, their

production is expensive and energy intensive, requiring petroleum feedstocks, which

contribute to greenhouse gas emissions, and harsh reaction conditions, especially

when making Group IV PAOs, whose synthesis requires high concentrations of

corrosive catalysts (AlCl3, BF3, and HF).18,21 To mitigate these challenges, promising

renewable alternatives from bioderived feedstocks have gained momentum.22–29

Bioderived feedstocks have unique functional groups that enable site-specific

chemistries during processing. In addition to platform chemicals, such as furfural,

derived from lignocellulosic biomass, fatty acids, obtained from natural oils, such as

coconut and palm kernel oils, can undergo ketonization to form long-chain ketones (24

6

+ carbon atoms). Ideal lubricant base oils contain anywhere between 18 to 40 carbon

atoms and consist of a long hydrocarbon backbone with a high level of branching to

enhance the oil’s physical properties, including viscosity index, pour point, and

oxidative stability.21

In one of the early attempts to make biobased products, Corma and co-workers

produced n-tricosane, a linear alkane product with 23 carbons, from lauric acid, a fatty

acid found in coconut and palm kernel oils.26 This process involved ketonization of

lauric acid to form 12-tricosanone, containing 23 carbons, followed by its

hydrodeoxygenation (HDO) to produce n-tricosane (58.2% selectivity) and C10

through C22 alkanes (13.9% selectivity total). Although the final product would not be

suitable as a Group III base oil owing to its poor viscosity and high melting point, the

short-chain alkanes may be a suitable for ultra-low sulfur diesel (ULSD) fuel. Even so,

USLD blends ($3.15 per gallon)30 are not as price competitive as Group III base oils

($5.01 per gallon).31 Despite these obstacles, the strategy proposed by Corma and co-

workers is appealing. The intermediate ketone, 12-tricosanone, can be obtained in

89% yield by ketonization of lauric acid (Figure 1-1) and is an ideal starting material

for lubricant synthesis because it contains a high number of carbon atoms and can

partake in carbon-carbon coupling reactions due to the presence of a ketonic group.

Ketones are excellent platforms to incorporate branching because of their

acidic -CH group at the α carbon position.28,29,32–35 Previously, Bell and co-workers

performed multiple cross-ketonization reactions, starting from short-chain length C3 –

C5 carboxylic acids, followed by HDO, to produce C12 – C33 branched and cyclic

alkane diesel fuels and lubricant base oils.27 Wang and co-workers produced C23

biolubricant base oil with ~50% yield from acetone and furfural via successive aldol

7

condensation and HDO steps.25 Although these routes use renewable feedstocks, for

example, carboxylic acid and acetone, which can be produced from sugar

fermentation, or lignocellulosic biomass-derived furfural, they require multiple

reaction steps and extractions. This results in carbon loss, the need for excessive

amounts of solvent, and high production costs. Accordingly, there is still a need to

provide novel strategies to synthesize a highly branched bio-lubricant base oil,

preferably using one or more bioderived starting material(s), for example a long chain

ketone obtained from bioderived fatty acids, and substituted or unsubstituted furfural,

obtained from lignocellulosic biomass.

1.3 Mild Hydrothermal Treatment of Catalysts Involved in Biomass Upgrading

This section focuses on mild hydrothermal treatment of catalysts. To put this

into context, well-ordered, siliceous materials called zeosils are commonly employed

as supports for metal catalysts in biomass upgrading, but as mentioned previously,

require frequent regeneration to remove carbonaceous by-products from their pores. It

is thought that studying the effects of mild hydrothermal treatment on silica-containing

materials will aid in understanding how regeneration methods affect the zeosil

framework and will provide information on how defect sites called silanol (SiOH)

groups may form in siliceous materials. Developing methods to enhance the number of

SiOH groups may be useful when aiming to incorporate more anchoring points for

metals into zeosils. The following two sub-sections provide more information on

hydrothermal treatment as it pertains to catalyst regeneration and the formation of

anchoring points for metals in silica-containing materials.

8

1.3.1 Hydrothermal Treatment to Regenerate Catalysts

Hydrothermal treatment involves steaming materials for an extended time at

high temperatures. In previous work, solid-state NMR experiments indicated that well-

ordered siliceous materials lose their crystallinity after being exposed to steam for

extended times, e.g., for 80 days at 823 K and 11 bar of steam.36 More recently, it was

shown that, even though changes related to steaming may not be evident using

traditional characterization methods, there may be important structural changes

affecting catalyst activity.37 Calcination is another method of regeneration, in which

carbonaceous byproducts react with air at elevated temperatures to form carbon

dioxide and trace amounts of steam. Steam at low concentrations is likely to interact

with silica’s surface, but these atomic-level interactions are challenging to detect by

methods used to study powders, such as solid-state NMR and XRD, which are better

suited to study the material’s bulk rather than its surface.38,39 These surface

interactions are important, however, as numerous technological applications of silica

rely on its specific surface properties. In particular, as mentioned previously, surface

silanol groups (SiOH) serve as anchoring points for a variety of chemical species.40

Surface science techniques, such as Infrared Reflection Absorption

Spectroscopy (IRRAS) and X-Ray Photoelectron Spectroscopy (XPS), offer

extraordinary atomic-level precision but work best when the surface is electrically

conductive. Unfortunately, siliceous materials are not electrically conductive, and

many of those techniques are not applicable to these materials.41 One way to overcome

this lack of conductivity is by casting or growing 2-D silicate thin films onto metal

single crystal substrates. An example of this is the bilayer silicate, which can be

prepared on various metal supports, but Ru(0001) has received the most attention.42–45

The bilayer silicate consists of corner-sharing tetrahedral [SiO4] building blocks

9

arranged in a honeycomb structure and can contain amorphous regions in addition to

the crystalline bilayer.44 Often times, polymorphous regions are commonly found in

metals supported on siliceous supports.46 The bilayer does not, however, contain the

wide range of 3-D pores and channels present in zeosils. Recently, all-Si MFI

nanosheets have been deposited onto metal substrates, especially onto Au(111), due to

its inert nature and thermodynamically stable (111) facet.47–49 MFI nanosheets contain

interconnected pore networks and have the same framework that makes up self-

pillared pentasil (SPP), an important zeosil for biomass upgrading reactions.50,51 The

deposition of 2-D siliceous materials onto conductive supports now provides a model

system that can be investigated through the lens of surface science. While many

surface science techniques have been traditionally limited to ultra-high vacuum

(UHV) conditions, technical developments in the last few decades have allowed

operation at pressures and temperatures more relevant to practical applications.52 This

then enables monitoring the effects of low concentrations of water vapor (e.g., 3 mbar)

at elevated temperatures (e.g., 573 K) on 2-D surfaces, in situ, using surface science

techniques, such as IRRAS. Understanding the effects of mild hydrothermal

conditions on 2-D siliceous nanomaterials may aid in understanding the effects of

these conditions on 3-D ones.

1.3.2 Hydrothermal Treatment to Anchor Metals

Phosphorus-containing zeosils (P-zeosils) have gained much attention in the

pursuit to convert biomass-derived platform chemicals to useful products.50,51 In

particular, P-BEA and P-SPP were found to convert biomass-derived dimethylfuran

(DMF) to para-xylene with 97% yield. These P-zeosils were more active and selective

than Al-BEA, Zr-BEA, and H3PO4.50 In another report, P-BEA, P-MFI, and P-SPP

10

were highly selective (>85%) in the conversion of tetrahydrofuran (THF) to

butadiene.51 Although P-zeosils are highly active in biomass chemistries, there is still

a lack of fundamental understanding regarding how phosphorus interacts with the

zeosil framework. Given the structural complexities associated with 3-D materials, 2-

D materials may be a better starting point to determine the interaction between

phosphorus and the silica-containing surface. However, 2-D materials often times lack

enough suspected anchoring points, such as SiOH groups, to interact with metal

species.

In the past, attempts have been developed to create anchoring points, or SiOH

groups, on 2-D surfaces. Prior studies have shown when water interacts with a silica-

containing framework, Si-O-Si linkages undergo hydroxylation to form SiOH.53–55 In

particular, Freud and co-workers found that only small amounts of SiOH groups could

be formed in a bilayer silicate upon deposition of an ice-like film of water at 100 K

and subsequent heating to room temperature.55 In a later study, Sauer and co-workers

found that the number of SiOH groups increased only when an ice-like water film on

the bilayer silicate was bombarded with electrons at cryogenic conditions and then

heated to room temperature.56 The formation of more SiOH groups by electron

stimulation came at the expense of partial, irreversible destruction to the bilayer’s

crystalline framework.56 It has yet to be determined if hydrothermal treatment can

create SiOH groups on 2-D surfaces, such as the bilayer silicate and MFI nanosheets.

1.4 Thesis Objective and Overview

Lignocellulosic biomass offers a promising alternative feedstock to petroleum,

but technological challenges prevent its commercialization. The objective of this thesis

is threefold: (1) to investigate the active species of the metal salt, AlCl3, in glucose

11

conversion to fructose, (2) to develop a catalytic process for the production of a bio-

lubricant base oil from renewable feedstocks, and (3) to provide a fundamental

understanding on the effects of mild hydrothermal conditions experienced during

catalyst regeneration on siliceous nanomaterials. Such knowledge will aid in the

development of improved catalysts and processes to obtain products from biomass.

This thesis consists of six chapters in addition to the Introduction.

Chapter 2 reviews metal salt hydrolysis, provides the experimental and

computational tools developed to study metal salt speciation, and describes the

catalytic role of metal salts in biomass upgrading.

Chapter 3 expands on the review of metal salt speciation provided in Chapter 2

by highlighting experiments involving AlCl3. Direct speciation measurements are

coupled with kinetics to provide strong evidence for the active species of AlCl3 in

glucose to fructose isomerization in water. The interplay between Lewis (AlCl3) and

Brønsted (HCl) acids is examined using a speciation model, while simultaneously an

experimental protocol is developed to quantify the various aluminum species. Linear

scaling between the glucose isomerization rate and the speciation measurements at

sufficiently high temperatures indicates that the hydrolyzed [Al(H2O)4(OH)2]1+

complex is the active species in glucose isomerization.

Chapter 4 delves into the production of a bio-lubricant base oil from renewable

feedstocks. The strategy involves coupling 12-tricosanone, obtained from bioderived

fatty acids, with furfural, a platform chemical obtained from lignocellulosic biomass,

to form a highly branched bio-lubricant base oil. The viscous properties of the final

product are comparable to commercial petroleum-derived Group III and Group IV

base oils.

12

Chapter 5 explores the effects of mild hydrothermal treatment on all-Si MFI

nanosheets and a polymorphous bilayer silicate. In the past, monitoring the effects of

mild hydrothermal conditions, in situ, on siliceous materials has been challenging to

observe by surface science methods, which often require electrically conductive

substrates. The emergence of 2-D siliceous nanomaterials deposited on metal single

crystals overcomes this limitation. In this work, elevated temperatures and pressures of

water increase the formation of SiOH groups in the MFI nanosheets, but do not change

the polymorphous bilayer silicate. The effects are fully reversible in the MFI

nanosheets. Implications shown here may be useful when considering mild

hydrothermal treatment on 3-D materials used in biomass upgrading.

Chapter 6 summarizes the key conclusions from the research conducted in this

thesis and addresses remaining challenges and future directions to advance the field.

13

HOMOGENEOUS METAL SALT SOLUTIONS FOR BIOMASS UPGRADING

AND OTHER SELECT ORGANIC REACTIONS

2.1 Introduction

Speciation, a term borrowed from the biological sciences, referring to the

various chemical forms of an element, has gained significant attention in chemistry.57

It has become especially important in environmental policy, occupational health,

nutrition, and medicine, where the characteristics of just one species of an element

may have a radical impact on living systems. An extreme example is organotin

compounds.58 Although these compounds are effective as pesticides, antifungal

agents, and marine anti-fouling agents, they result in serious harm to the endocrine

system.59 Furthermore, they are only a minute part of the total amount of tin used in

such applications; the remainder being inert tin oxides.60 Measuring these individual

organotin compounds is therefore critical, as these species are the main threat to living

organisms.

In catalysis, an analogous case to that of tin occurs when we aim to study the

active species of metal salts as catalysts. Metal salts form multiple species upon

dissolution in aqueous media, yet determining the active species is challenging. The

concentrations of chemical species vary depending upon changes in temperature,

pressure, pH, humidity, presence of organic matter, etc.61 When developing techniques

to identify species, the following questions often arise:57 (1) What are the species we

want to measure? (2) How should we sample the material and/or isolate the species

Chapter 2

14

without changing its composition? (3) Can we detect very low amounts of isolated

species, which may represent only a minute fraction of the total, already ultra-trace

element concentration? In this section, we aim to answer these questions by presenting

an overview of metal salt hydrolysis and describing the experimental and

computational tools developed to study metal salt speciation. We then describe the

catalytic role of metal salts in biomass upgrading.

2.2 Metal Salt Hydrolysis

Metal salt hydrolysis is the reaction of metal cations with water to liberate

protons and form hydroxy or oxy complexes in solution.61 In aqueous media, metal

salts dissociate to form metal (M) cations. These cations are solvated by water,

forming complexes, M(H2O)nx+, which undergo hydrolysis, as shown in Eq. 2-162 or

Eq. 2-261 (when neglecting the hydration sphere of water, which is not always known):

M(H2O)nx+ + y H2O ↔ M(H2O)n-y(OH)y

(x-y)+ + y H3O+ Eq. 2-1

Mx+ + y H2O ↔ M(OH)y(x-y)+ + y H+ Eq. 2-2

15

Highly charged, small ionic radius metal cations undergo the greatest degree of

hydrolysis due to their strong interaction with the surrounding lone pairs of the

oxygens in water. In turn, the hydrogens in water become more positively charged and

the bond between O-H weakens to produce protons.61 Figure 2-1 depicts the

hydrolysis mechanism and the products, i.e., a Lewis acid species ([M(H2O)n-

y(OH)y](x-y)+) and a Brønsted acid species (H3O

+); both are important when

determining the role of metal salts in biomass upgrading, as shown later in this review.

Eq. 2-3 is the general equilibrium quotient (QA):61

QA

=[M(OH)y

(x-y)+][H+]y

[Mx+][H2O]y Eq. 2-3

A large value of QA indicates the metal salt readily undergoes hydrolysis. At

low concentrations, the metal ion exists predominately as a mononuclear (one metal

ion) hydroxy species. As the pH increases, additional hydroxide ligands form, up to a

total that often reaches four or more:63

Mx+ → MOH(x-1)+ → M(OH)2(x-2)+ → M(OH)3

(x-3)+ → M(OH)y(x-y)+

Figure 2-1: First step in metal salt hydrolysis for a trivalent cation [M(H2O)6]3+.

16

Polynuclear (more than one metal ion) and solid species may form depending

on the pH and properties of the metal ion. Polynucleation occurs slowly and may

result in a variety of species. After long periods of time, and accelerated in basic

conditions, the polynuclear species form amorphous and eventually crystalline

solids.63 When studying metal salts as catalysts, although conditions are usually acidic,

polymeric species and solids may form over time and therefore should not be

overlooked. This speciation is critical to determining the active species and its

concentration for catalysis. Methods to achieve this are described next.

2.3 Measurements and Modeling of Elemental Speciation

2.3.1 Experimental Tools

Metal salt hydrolysis alters the speciation; quantifying the species and their

concentrations is an important research goal. Here, we review and discuss select

techniques to study speciation.

Electrospray ionization mass spectrometry (ESI-MS) is a soft ionization

technique that transfers pre-existing ions in solution to the gas phase and works best

for dilute solutions.16 ESI-MS yields the mass to charge ratio (m/z) and information on

the oxidation state of a metal ion, but should not be used to identify the structure of

metal complexes. This is because during the transition of charged droplets to the gas

phase, water molecules surrounding a metal center evaporate until the existing

solvation sphere is no longer able to stabilize the charge on the metal ion center. This

in turn reduces the overall charge on the metal ion complex through the loss (or

ionization) of the water ligands. The overall process is equivalent to a gas-phase

hydrolysis reaction.16 In addition, electrochemical processes at the electrospray

17

electrode can induce changes in the charge state and lead to alteration of the species in

solution.64 There are reports involving aqueous metal salt solutions where ESI-MS

detects only [M(H2O)n(OH)2]1+, n=0 – 4.65–68 For example, Urabe et al. only observed

monomeric [Al(H2O)n(OH)2]1+ species at any concentration of AlCl3 (0.02 – 100 mM).

This result is puzzling because (1) given the pH of their salt solutions (pH<3), the

hexa-aqua species is expected to predominate and (2) there were no differences

between the ESI-MS spectra at varying concentrations of AlCl3.65 This coincidence is

most likely because the actual species have been altered during the electrospray

ionization process.

Inductively coupled plasma mass spectrometry (ICP-MS) detects metals down

to ultra-trace concentrations, but cannot resolve individual metal species, since all

molecules are broken down when exposed to the high temperature ion source.57 It is

frequently used to determine if metals have leached from heterogeneous catalysts and

confirm the concentrations of metal salt catalysts.

Nuclear magnetic resonance (NMR) had until recently limited quantitative

accuracy. Today, NMR is quantitative up to three significant figures, is non-

destructive, and applies only a small energy perturbation to a system using a low-

energy electromagnetic wave. Furthermore, in situ analysis of reacting systems is

possible when the free energy change of the reaction is small and the reaction rate is

slow.69–71 When studying metal salt speciation, Maki et al. developed a method to

quantify the [Al(H2O)6]3+ species by creating a calibration curve with varying amounts

of aqueous Al(NO3)3 in HNO3.72 The calibration showed excellent linearity for the

[Al(H2O)6]3+ species, from 1×10-4 to 1 M. While the technique worked well for

18

aluminum speciation, it has not been applied to other trivalent metals, as the nuclei are

not as common and the signals are often broad.73

Ultraviolet-visible spectroscopy (UV-Vis) detects metal complexes that absorb

light at different wavelengths. In 1957, Elving et al. studied the coordination chemistry

of CrCl3 in aqueous solutions of HCl using UV-Vis.74 Wavelengths were observed for

species corresponding to the [Cr(H2O)6]3+, [Cr(H2O)5Cl]2+, [Cr(H2O)4Cl2]

1+, and

[Cr(H2O)3Cl3] complexes, where the chlorinated complexes appeared at increasing

wavelengths. Recently, others have performed similar studies with CrCl3 in ionic

liquids,75 highly concentrated LiCl,76 and choline chloride (CrCl3/xH2O/yChCl),77

which corroborate Elving et al.’s finding. To the best of our knowledge, the chromium

species have not been quantified through this method. UV-Vis is limited to solutions

that absorb light within the UV range (190 – 380 nm) and the visible light range (380

– 750 nm). Therefore, clear solutions, such as those containing aluminum salts, are not

well suited for this technique.

X-ray absorption spectroscopy (XAS) measures the transitions from the core

electronic states of the metal (X-ray absorption near-edge structure, XANES) to the

excited electronic states and the continuum (extended X-ray absorption fine structure,

EXAFS).78 XANES reports the electronic structure and symmetry of the metal site,

while EXAFS reveals the type of nearest neighbors, their distances from the metal

core, the coordination number, and the Debye-Waller factors, which are indicative of

the degree of vibrational and static disorder.57 Brown et al. combined XANES and

EXAFS to study the structure and speciation of complexes for Fe2+ and Fe3+ chloride

solutions as a function of pH, ionic strength, and chloride/ion ratios. XANES revealed

that less acidic solutions contain octahedral complexes, while acidic FeCl3 solutions

19

exist as tetrahedral complexes. For acidic solutions, EXAFS showed an increase in the

first-shell Fe3+-ligand distance, indicating the octahedral complex changes (Fe3+-O

bond distance of 2.10 Å) to a tetra-chloro complex (Fe3+-Cl bond distance of 2.23 Å).

FeCl2 solutions remained octahedral, but one of the ligands surrounding Fe2+ had

exchanged a water for a chloride ion. This study provided more direct evidence of Fe2+

and Fe3+ structural complexes compared to spectroscopic studies, which rely on the

assignment of absorption bands to particular species that are not always known for

metal species.79 In a recent study with CrCl3, Verdonck et al. performed EXAFS to

identify complexes observed at different spectral bands in UV-Vis.76 EXAFS revealed

each band corresponded to a different chromium species. EXAFS also detects catalyst-

reactant interactions. For example, EXAFS analysis by Choudhary et al. detected the

presence of a glucose molecule in the first coordination shell of the Cr cation,

indicating Cr assists in ring-opening and isomerization of glucose.7 While XAS is

powerful, data analysis is challenging, and, when performing EXAFS, it is often

difficult to distinguish between scattering atoms with similar atomic numbers (C, N, O

or S, Cl or Mn, Fe).78

Small angle neutron (SANS) and X-ray (SAXS) scattering as well as static

(SLS) and dynamic light (DLS) scattering apply radiation to a sample, causing the

particles in solution to scatter light and produce a scattering pattern to yield

information on the particles’ size, shape, and orientation.80 Light scattering techniques

are suitable for dilute particle solutions with a radius of gyration (Rg) between 10 and

1000 nm, while SAXS and SANS are suitable for structures on a smaller scale (as low

as 0.1 nm), as well as for probing particle internal structures and in highly

concentrated samples.81 Shafran et al. traced the size of soluble polymeric aluminum

20

species by DLS and found that particle size increases over time and in basic

conditions. These results supported the hypothesis that Al undergoes polymerization

from an Al13-mer to Al30-mer.82 In another study, Choudhary et al. explained that the

formation of Cr particles detected by DLS cause the pH of CrCl3 solutions to drop and

remain stable upon heating.7 Recently, SAXS determined the size and shape of SnO2

nanoparticles formed from SnCl4.83 SAS possesses low resolution and is unable to

distinguish closely related species (e.g., monomer and dimer)80 but can provide

valuable information on the formation of particulates that reduce the concentration of

the active species for catalysis.

2.3.2 Computational Tools

Models based on thermodynamic equilibria can quantify speciation. In our

experience, the OLI Systems, Inc. “Stream Analyzer” software (2018) is a powerful

modeling platforms for this purpose. It predicts metal salt speciation by considering

chemical equilibria of multicomponent systems and combining standard-state

thermochemical properties of solution species with an expression for the excess Gibbs

free energy.84–88 The standard state properties are calculated using the Helgeson-

Kirkham-Flowers (HFK) equation of state for aqueous ions and electrolytes at infinite

dilution, while the excess Gibbs free energy model incorporates short, middle, and

long-range electrostatic interactions. We have seen agreement between OLI’s

predictions and our experimental data for AlCl3 salt solutions.89 Furthermore, OLI has

established its model can reproduce the pH, solid solubilities, vapor-liquid equilibria,

and other properties for metal species in pure water and organic components. 84–88

The OLI software allows the user to select either an aqueous (AQ) or mixed

solvent electrolyte (MSE) framework to model speciation. The AQ framework allows

21

for modeling 80+ elements in the range of 223 to 573 K, 0 and 1500 bar, and up to 30

molal ionic strength. The MSE framework contains about half the species as the AQ

databank, yet provides complex chemistries that cannot be simulated using the AQ

framework. The MSE framework is recommended if the chemistry of interest is

available.90 Figure 2-2 is an example of the OLI software’s predictions on the effect of

pH on the dominant ionic species of AlCl3. The main species include the hexa-aqua

species, Al3+, the stable, crystalline solid form of aluminum hydroxide, called

boehmite, AlO(OH), and the mono- and di-hydroxy species, Al(OH)2+ and Al(OH)21+,

respectively. Al3+ dominates at high acid concentrations.7,91 In addition, the prediction

of solid boehmite coincides with Hem’s findings, which demonstrated that heating

aqueous aluminum results in the formation of boehmite, whose structure was

confirmed by x-ray diffraction (XRD).92 Finally, Mesmer et al. determined the

hydrolyzed species are most significant at relatively low concentrations of aluminum

(e.g., ≤10 mM), consistent with OLI predictions,93 and Norton et al.89 recently

observed qualitative agreement between OLI predictions and experimentally deduced

concentrations of the hydrolyzed aluminum species. In summary, the OLI software’s

predictions are consistent with the hydrolysis behavior of metal salts and experimental

findings.

22

Although our experience has mainly been with the OLI speciation model, other

platforms exist for modeling chemical speciation. We discuss two additional, open-

access speciation models for the reader’s reference. The first is called the Visual

MINTEQ, which can simulate the chemical composition in solutions in contact with

gases, solid compounds, and particle surfaces. It calculates the speciation of inorganic

ions and complexes in water and determines the conditions that result in precipitation

of solids.94 Compared to OLI, Visual MINTEQ only works for a limited number of

salts and conditions (e.g., most accurate between 273 and 313 K and near 1 atm). The

second model is a multi-purpose chemical speciation program called Geochem-EZ.95

It allows the user to perform equilibrium speciation computations, estimate solution

ion activities and pH, and determine interactions between metals and ligands. The

program is capable of modeling chemical speciation in soils, but, as with the Visual

MINTEQ, it is rather limited in the number of systems for which databases exist

compared to OLI.

0 10 20 30 40 500

1

2

3

4

5

AlO(OH)(solid)

Al3+

HCl (mM)

0.00

0.02

0.04

0.06

0.08

0.10

Al(OH)21+

AlOH2+

AlO

(OH

) (so

lid

), A

l3+

(mM

)

Al(

OH

) 21+, A

lOH

2+

(mM

)


(413 K), calculated using the OLI speciation model. The hydration sphere of water has

been removed for clarity. Redrawn from Norton et al.89

23

Quantum chemistry based thermodynamic models, such as COSMO-RS

( COnductor like Screening MOdel for Real Solvents), provided the use of the

appropriate electrolyte extensions, can be powerful tools for the assessment of the

effect of metal salts in the thermodynamic properties of the solution, particularly in

biphasic systems.96 These effects become critical, as even small concentrations of salts

in solution can have a significant impact on phase equilibria and mutual solubilities of

solvents, all important properties for purification processes.97

Extensions to the current model such as those developed by Wang et al.98 and

Hsieh et al.99 have enabled the prediction of mean ionic activity coefficients in

electrolyte solutions by adding the Pitzer Debye–Hückel term to account for long-

range interactions and was successfully applied for few representative sodium salts.

Further advances by Ingram et al. enabled the use of COSMO based models for

solutions containing monoatomic fully dissociated salts.100 They introduced an

element specific COSMO-radii for alkali metals that consider cation hydration and

element specific hydrogen bonding contribution for anions. This extension proved

successful at: predicting mean ionic activity coefficients in mixed solvent systems,

describing salt effects on different liquid-liquid equilibria, predicting salt induced

separation of aqueous and organic systems, and calculating the partition coefficients of

organic solutes, such as HMF, into different organic solvents in the presence of metal

salts.100,101

Lastly, electronic structure calculations, such as ab initio molecular dynamics

(AIMD) and density functional theory (DFT) calculations, provide mechanistic

information on elementary reactions and (free) energy barriers that can complement

experimental work. For example, Mushrif et al. investigated the mechanism of glucose

24

ring opening and isomerization to fructose catalyzed by CrCl3 in the presence of water

using Car-Parrinello AIMD calculations while accounting for solvent dynamics at

finite temperature.102 Energy barriers were much lower for [Cr(H2O)5(OH)]2+

compared to the hexa-aqua Cr3+ species, suggesting [Cr(H2O)5(OH)]2+ is the active

species. These simulation results are in agreement with experimental findings, as will

be addressed later in this Review.

2.4 Metal Salts in Biomass Upgrading

Monosugars, such as glucose and xylose, from the cellulosic and hemi-

cellulosic parts of biomass, are platforms for chemicals, such as lactic acid, HMF,

levulinic acid, furfural, and xylitol (Figure 2-3).103

Figure 2-3: Glucose and xylose as platform chemicals with select products shown.

25

Lactic acid and HMF syntheses happen via hydrolysis of cellulosic biomass to

glucose, glucose isomerization to fructose, and fructose dehydration to either lactic

acid or HMF. The product distribution depends upon the catalyst and reaction

conditions (Figure 2-4).12 In this section, we discuss the role of metal salts in the direct

conversion of cellulose to lactic acid. Subsequently, we address the tandem catalysis

of glucose to HMF and identify methods to detect the active species responsible for

glucose isomerization to fructose. Finally, we recommend future work.

Lactic acid has received much attention as a monomer for the production of

biodegradable plastics. Currently, it is produced by fermentation of glucose; however,

this biological process cannot be directly applied to cellulose, and cellulose must first

undergo hydrolysis. In an effort to combine multiple reaction steps in a single pot,

Wang et al. found that lactic acid can be directly produced from cellulose with 68%

yield using dilute metal Pb2+ in water.104 Nitrates of Al3+, Bi3+, In3+, and Zn2+ were

also tested, but resulted in low yields of lactic acid (<20%). The results have inspired

research of other catalytic systems for the direct conversion of cellulose to platform

chemicals. For example, Tang et al. determined that VOSO4, a less harmful compound

compared to Pb2+, could catalyze the transformations of glucose and cellulose into

Figure 2-4: Cellulose hydrolysis followed by glucose isomerization to fructose and its

subsequent transformation to HMF and lactic acid.

26

either formic acid or lactic acid by simply changing the reaction atmosphere from O2

to N2.105 VOSO4 exhibited even higher yield to lactic acid under O2 than Pb2+ does.

51V NMR spectroscopy and electron spin resonance (ESR) suggested that VO2+ was

the active species responsible for converting the intermediate glucose into lactic acid.

In another study, lanthanide triflates were found to convert cellulose into lactic acid.106

Yields as high as 89.6% were obtained with Er(OTf)3. ErCl3, a less expensive

alternative, also resulted in high yield of lactic acid (91.1%). In summary, these

studies suggest that metal cations play an important role in the conversion of cellulose

to lactic acid.107

In addition to lactic acid formation, recent efforts have focused on the

conversion of glucose to HMF, a platform chemical for biofuel, biochemical, and

biopolymer industries. A bottleneck in the production of HMF is the glucose

isomerization, an equilibrium limited reaction. Currently, industrial practice uses

immobilized D-xylose ketoisomerase to catalyze the isomerization;108 however, these

catalysts are expensive, operate under narrow pH conditions, and require highly pure

glucose.109 One way to overcome the equilibrium limitations is to combine the

isomerization and dehydration steps in a single pot. This tandem reaction scheme

requires a catalyst that is compatible with the Brønsted acidity and high temperatures

associated with the dehydration reaction. As a result, over the past decade, research

has turned to the development of chemo-catalysts to carry out glucose isomerization.

Following the pioneering work of Davis and co-workers,11 who combined

heterogeneous Lewis acid catalysts, e.g., Sn-BEA, with strong inorganic Brønsted

acids, e.g., HCl, metal salts were exploited as homogeneous Lewis acid catalysts for

isomerization. In one of the first studies involving metal salts, Heeres and co-workers

27

found that Al3+, Cr2+, and Zn2+ resulted in the highest conversion of glucose to either

HMF or lactic acid.110 Subsequently, Zhang and co-workers discovered that CrCl3 in

ionic liquid performed tandem catalysis of the isomerization as well as dehydration

chemistry,13 and this was originally puzzling. Following this work, it was proposed,

consistent with the hydrolysis reaction (Eq. 2-3), that metal salts function as Lewis

acids (metal species) and Brønsted acids (hydronium ion).111 However, because

acidification of the solution is not as strong (it depends on metal and salt

concentration), Brønsted acids are added to metal salts to further acidify the solution,

speed up the dehydration chemistry, and promote the selective formation of HMF

(Figure 2-4).

Pagán-Torres et al. studied the conversion of glucose to HMF with HCl and

either AlCl3, SnCl4, VCl3, InCl3, GaCl3, LaCl3, DyCl3, or YbCl3. The results indicated

that the effectiveness of the metal salts in glucose conversion depended upon the salt’s

Lewis acid hardness and the size of the ionic radius.112 Hard Lewis acids interact with

hard Lewis bases. AlCl3, the hardest of the Lewis acids studied, is the most active and

interacts strongly with the hard Lewis bases in the hydroxyl groups in glucose.

Similarly, the size of the salt’s ionic radius correlates to its reactivity of glucose.

Glucose conversion increases with decreasing ionic radii from In3+ > Ga3+ > Al3+ and

for the lanthanide series from La3+ > Dy3+ > Yb3+. This behavior is most likely due to

the strong electrostatic interaction between glucose and the smaller cations.

In another screening study by Enslow et al., SnCl4 was identified as the most

selective Lewis acid for converting glucose to HMF (yield < 17%) in water.113 The

presence of HMF was an indication SnCl4 underwent hydrolysis to form complexes,

namely the octahedrally coordinated aquachlorotin(IV), [Sn(H2O)6-xClx](4-x)+ and/or

28

the hexaaquatin(IV), [Sn(H2O)6-y(OH)y](4-y)+, to drive glucose isomerization and

protons to drive fructose dehydration. 119Sn-NMR showed low concentrations of

SnCl4 (<100 mM) led to only a single peak corresponding to hexaaquatin(IV) ions,

suggesting chloride-containing complexes do not form. In addition, water-insoluble

SnO2 species were observed after SnCl4 solutions (10 – 60 mM) were kept at room

temperature for 24 hr. The exact distribution on SnO2 in sub-100 mM SnCl4 solutions

was unknown, but these species in combination with the [Sn(H2O)6-y(OH)y](4-y)+ were

thought to contribute to the Lewis acidity necessary for glucose isomerization.

Similarly, Barros dos Santos et al. found that Sn4+ containing homogeneous catalysts,

including butylstannoic acid (BTA), di-n-butyl-oxo-stannane (DBTO), and dibutyltin

dilaurate (DBTDL), were active in converting cellulose to glucose, fructose, lactic

acid, and HMF.114 Each catalyst resulted in similar cellulose consumption and yield. In

an effort to determine the active species, SnO2, a species common to all of the tested

Sn4+ containing catalysts, was also studied and gave comparable conversion and yield

as the other tested Sn4+ catalysts. This suggests an oxide was the active Sn4+ species,

consistent with Enslow et al.’s proposition when using SnCl4 as a catalyst to make

HMF from glucose.114

While SnCl4 is selective to HMF in water, CrCl3 and AlCl3 are highly active in

glucose isomerization to fructose. To better understand the differences in reactivity

between the Lewis acids, Tsilomelekis and co-workers studied the effects of the metal

salts on the ease of ring opening of glucose, the first step in glucose isomerization.115

At the same metal salt concentration, it was found that SnCl4 facilitates fast

mutarotation towards the β-anomer, while CrCl3 and AlCl3 favor the α-anomer. AlCl3

was the slowest on glucose ring opening, though this effect is highly dependent on pH.

29

For example, when the pH is ~2.9 (50 mM CrCl3 and 100 mM AlCl3), AlCl3 facilitates

mutarotation much faster compared to CrCl3, suggesting the induced Brønsted acidity

may play a role in glucose mutarotation. Overall, these findings are consistent with

studies on the anomeric specificity of enzymes, which show that immobilized d-

glucose isomerase has 110% higher conversion rate starting from the α-anomer as

compared to the β-anomer,116 and could explain differences in reactivity among the

metal salts.

Following glucose ring-opening is the intramolecular hydride shift from the C2

to the C1 position of glucose. Choudhary et al. showed that the mechanism of glucose

to fructose in CrCl3 and AlCl3 is analogous to that of heterogeneous Sn-BEA zeolite

(Figure 2-5).14 It was proposed that the active species for both homogeneous and

heterogeneous catalysts likely exist as a bifunctional Lewis-acidic/Brønsted-basic site,

in which the Brønsted base (-OH) facilitates an initial proton transfer, activating the

subsequent Lewis acid (metal center, M) catalyzed C2→C1 hydride shift.102 Similarly,

theoretical computations were performed to better understand the catalytically active

species for glucose to fructose isomerization when catalyzed by Pb(NO3)2. When

comparing the [Pb(OH)]1+ species to the Pb2+ species, it was found that [Pb(OH)]1+

significantly lowered the activation barrier for the C2→C1 hydride shift. The

mechanistic similarities observed between homogeneous and heterogeneous catalysts

suggest that knowledge from homogeneous catalysts, which are easier to study and do

30

not invoke synthesis challenges, could be applied to the identification and

improvement of heterogeneous catalysts.

Designing better catalysts requires understanding the active species responsible

for the chemistry. Different approaches have emerged to identify the catalytically

active species of homogeneous metal salts in the glucose isomerization reaction.

Choudhary et al. combined speciation modeling using the OLI software and glucose

isomerization kinetics to propose an active species of CrCl3.7 The initial rate increased

linearly with the concentration of the [Cr(H2O)5(OH)]2+ species, suggesting this is the

most active species. Furthermore, Car-Parrinello molecular dynamics (CPMD)

simulations and EXAFS confirmed that Cr3+ exists as a hexahydrate species,

surrounded by six oxygens in the absence of glucose. Upon glucose addition, two

oxygen atoms of the hydroxyl groups in glucose replace two of the oxygen molecules

in the inner coordination sphere. The metal center is then coordinated to three water

molecules, one OH-, and two oxygen atoms of the glucose molecule. This

complexation of glucose with the active [Cr(H2O)5OH]2+ species is critical in the

isomerization of glucose to fructose.62

While active species have been proposed, direct experimental evidence in

support of the catalytic species has been lacking. This lack of understanding stems

1

2

D-glucose D-fructose

2

1

1

2

Deprotonation

transition state

Intermediate-1 Hydride shift

transition state

Reprotonation

transition state

Intermediate-2


represents a metal center [e.g., Cr, Al, and Sn (in Sn-BEA)]. Redrawn from

Choudhary et al.14

31

from (1) a number of metal hydrolysis steps and self-condensation leading to

oligomeric species and potentially metal nanoparticles and (2) the concentration of

metal salts being low (e.g., 5 – 30 mM), making it difficult to observe and quantify the

metallic species. Despite these challenges, experimental studies have been performed

to elucidate the interaction between the active metal species and glucose. Tang et al.

observed a complex corresponding to [Al(OH)2(Glucose)]1+ (m/z=385) using ESI-

MS/MS15 and proposed [Al(OH)2]1+ is active for glucose isomerization. This finding

was in line with mechanistic results by Qi et al., which showed glucose isomerization

with AlCl3 proceeds with either [Al(H2O)4(OH)]2+ or [Al(H2O)2(OH)2]1+ as the active

species;117 however, this is not conclusive, as ESI-MS may alter the species’ structure.

Similar studies have been performed with metal salts that have moderate Lewis

acid strength (pH ~ 5 – 7). For example, Cu(NO3)2 was recently studied for glucose to

fructose isomerization, and the influence of pH was carefully monitored before and

after the reaction. Depending on the pH value, the thermodynamic equilibrium shifts

towards either Cu2+ or Cu(OH)1+ and Cu(OH)2. Correlation of the fructose yield with

the concentration of Cu(OH)1+ suggested this was the active Lewis acidic species.

According to the thermodynamics, the highest concentration of Cu(OH)1+ was

achieved at pH 5.3 – 5.5, which also provided the highest fructose yield of 16% at 383

K. ESI-MS was employed to investigate the interactions between different Cu species

and glucose. The resulting spectrum of glucose and Cu(NO3)2 revealed the presence of

the single-charged ion complex. FeCl3, another metal salt with moderate Lewis acid

strength, was also studied in glucose isomerization to fructose.68 Fe has two major

oxidation states, Fe2+ or Fe3+. Up until this study, the active species was expected to be

32

Fe3+. A combination of mass spectrometry, UV-Vis, and XAS showed that Fe3+ is

readily reduced to Fe2+ (over 95%) in the early stage of carbohydrate conversion.

Thus far, we have presented cases where metal species have been postulated,

but never experimentally quantified, as this is challenging. Norton et al. showed that

Al species’ concentrations can be deduced by combining experiments, modeling, and

glucose isomerization kinetics.89 In that work, 27Al-NMR detected the dominant

Al(H2O)63+ complex. These measurements combined with pH and equilibrium

expressions for the aluminum hydrolysis reactions were used to estimate the

concentrations of the [Al(H2O)4(OH)2]1+ and [Al(H2O)5(OH)]2+ species. DLS detected

aluminum particles upon heating the salt solutions. Over time and in the presence of

heat, it was suggested the polymeric Al species undergo nucleation to form stable

solid precipitates, such as boehmite. Solid aluminum was then quantified by coupling

ultrafiltration with ICP-MS. It was found that after heating catalyst solutions for 24 h,

the solid species and pH remained constant. Therefore, to ensure equilibrium of

aluminum species and avoid speciation changes during kinetic experiments, catalyst

solutions were preheated for 24 h prior to performing glucose isomerization. Finally,

the concentrations of the aqueous and solid aluminum species were correlated to the

glucose isomerization rate. At sufficiently high temperatures (e.g., 413 K), the

hydrolyzed [Al(H2O)4(OH)2]1+ species scales linearly with the glucose isomerization

rate, indicating this complex is the active one. This finding supports the idea that the

catalytic species responsible for the isomerization exists as a bifunctional Lewis-

acidic/Brønsted-basic site and provides additional support for the Al(OH)21+ active

species proposed by Tang et al.15

33

In summary of the published studies, it is evident that metal salts are crucial in

glucose conversion. They affect glucose ring opening, and their speciation facilitates

the C2→C1 hydride transfer to produce fructose, but detecting the active metal species

responsible for glucose conversion is challenging. When considering future

experiments, it is important to allow time for the metal species to equilibrate to

prevent changes in metal speciation during glucose conversion. The time for

equilibrating the salt species depends on the metal salt and temperature, and a simple

way to estimate it is by measuring the pH vs. time, prior to adding reagents (other

speciation signatures, such as particulates, etc. should also be monitored if possible;

long enough equilibration should be sufficient). While preheating equilibrates the

metal species, future studies are necessary to understand how preheating affects the

reaction rate and distribution of products.

Given the complexities associated with measuring speciation, models, such as

the OLI software, should be used to model pH and metal speciation and serve as a

starting point for future experiments. For example, when looking at the OLI

calculations for SnCl4 (Figure A-1), CrCl3 (Figure A-2), and AlCl3 (Figure 2-6) at

kinetically relevant glucose reaction conditions (413 K and 5 mM salt concentration),

the acidities of all the metal salt solutions are similar as a function of HCl

concentration (ranging from pH ~2.25 to 0.9, Figure 2-6a). The pH values for CrCl3

and AlCl3 eventually collapse onto the pH measured for HCl, indicating that at high

acid concentrations, the change in pH for these metal salt solutions is mainly dictated

by the HCl and not hydrolysis of the salt. Furthermore, CrCl3 and AlCl3 display a

maximum in pH at HCl concentrations of 1.5 and 17.9 mM, respectively (Figure

2-6a). The observed maximum in pH coincides with the observed maximum in the

34

proposed active species. High concentrations of Brønsted acids dissolve the solid to

increase the concentration of active species but also retard hydrolysis (Eq. 2-1) that

produces the active species; this tradeoff creates a volcano. This volcano-like behavior

was observed experimentally for the Al(OH)21+ species by Norton et al.89 It is shown

as the data points in Figure 2-6.

In contrast, for SnCl4, the pH continues to drop upon increasing HCl

concentration, and the speciation profile shows that solid formation proliferates upon

the addition of acid (Figure 2-6, Figure A-1), consistent with previous reports of solid

formation.83 Sn4+ is also more acidic compared to the other metal salts and HCl. This

may be because SnCl4 has a higher propensity to deprotonate water, always resulting

in a lower acidity compared to the other metal salts and HCl. Figure A-3 shows that

increasing the SnCl4 concentration results in high amounts of SnO2(solid) formation. All

the results combined, it is important to consider in future experiments, the relationship

between pH, the maximum active species concentration, and the maximum yield of

fructose, as well as the role of solids, such as SnO2, in glucose conversion.

The maximum active species concentration increases modestly with increasing

salt concentration (Figure A-4 and Figure A-5). Given that solids form at low external

acid concentration, increasing the rate beyond that corresponding to the volcano

maximum would not be possible. Anchoring the actives species on a support to

prevent agglomeration may be worth pursuing in future work. While analogy to

heterogeneous catalysts has been drawn, more work is needed to fully delineate the

similarities to and differences from heterogeneous catalysts. Furthermore, the number

of ab initio molecular dynamics and classical molecular dynamics calculations for this

class of catalysts has been rather limited. These calculations can provide further

35

insights into mechanisms and solvation and complement the bulk thermodynamic

models, such as OLI. Classic DFT methods are rather limited for charged systems and

solvent/substrate dynamics.

Figure 2-6: (a) Effect of HCl on pH for metal chlorides. The intrinsic (self) acidity of

solution (without HCl addition) is a function of the metal. (b) Proposed active species

of Al(III) and Cr(III) chlorides as a function of HCl concentration. Data points were

experimentally deduced Al(OH)21+ from Norton et al.89 The volcano curve is clearly

seen even though quantitative agreement with the OLI predictions is not great. (c)

Proposed active species of Sn(IV) as a function of HCl concentration. Salt

concentration is 5 mM and temperature is 413 K.

0 10 20 30 40 50

4.99982

4.99983

4.99984

4.99985

HCl (mM)

0 10 20 30 40 50

1.25

1.50

1.75

2.00

2.25

2.50

pH

HCl (mM)

5 mM AlCl3

5 mM CrCl3

5 mM SnCl4

HCl

(a) (b)

(c)

0 10 20 30 40 50

0.003

0.006

0.009

0.012

0.015

0.018

0.00

0.25

0.50

0.75

1.00

1.25

1.50

HCl (mM)

Al(

OH

) 21

+(m

M)

Cr(

OH

)2+

(mM

)

Al(OH)21+

(experimental)

Sn

O2

(so

lid

)(m

M)

36

DIRECT SPECIATION METHODS TO QUANTIFY CATALYTICALLY

ACTIVE SPECIES OF AlCl3 IN GLUCOSE ISOMERIZATION

3.1 Introduction

The need to reduce greenhouse gas emissions and our dependence on fossil

fuels has prompted considerable research in the production of fuels and chemicals

from lignocellulosic biomass.12 Biomass is a promising renewable feedstock due to its

abundance and its ability to capture CO2 from the atmosphere via photosynthesis.118

Among the top biomass-derived chemicals, 5-(hydroxymethyl) furfural (HMF) has

gained significant interest as a platform chemical.118 HMF synthesis results from the

hydrolysis of cellulosic biomass to glucose, glucose isomerization to fructose, and

fructose dehydration to HMF.12 When the isomerization and dehydration steps occur

in a single pot, dehydration drives the equilibrium-limited isomerization resulting in

higher yields.14 This then requires an isomerization catalyst that is compatible with the

Brønsted acid catalyst and high temperatures associated with the dehydration reaction.

Currently, industrial applications use immobilized D-xylose ketoisomerase to catalyze

the isomerization;108,119 however, these catalysts are expensive, operate under narrow

conditions, and require highly pure glucose.120 As a result, research has turned to the

development of chemo-catalytic processes to carry out glucose isomerization, starting

with the pioneering work of Davis and co-workers on Sn-BEA zeolite that enable

HMF production in a single pot.11,121

Chapter 3

37

Metal halides are related homogeneous chemo-catalysts.14,15,112,122,123 Hu and

co-workers compared the catalytic activity of AlCl3, CrCl3, and SnCl4 and found that

CrCl3 exhibits the highest initial turnover frequency (TOF), while AlCl3 is the most

selective catalyst to fructose.15 Mechanistic studies of glucose isomerization with

AlCl3 and CrCl3 show that the isomerization proceeds through a 1,2 hydride transfer,

and this step is rate-limiting according to kinetic-isotope effect measurements.14 Based

on theoretical investigations,124 the catalytic species responsible for the isomerization

likely exists as a bifunctional Lewis acidic/Brønsted base site, in which the Brønsted

base (-OH) facilitates an initial proton transfer, activating the subsequent Lewis acid

(metal center, M) catalyzed C2→C1 hydride shift (Figure 3-1).122 Direct experimental

evidence in support of the catalytic species has though been lacking. This lack of

understanding stems from a number of metal hydrolysis steps and self-condensation

leading to oligomeric species and potentially metal nanoparticles.125–127

Different approaches have emerged to identify the catalytically active species.

Our group combined kinetic rate measurements with a speciation thermodynamic

model, called Optimum Logic Inc. Systems’ Stream Analyzer software (OLI software,

2017), to suggest [Cr(H2O)5(OH)]2+ is active in CrCl3–catalyzed glucose

isomerization.7 Hu and co-workers performed tandem electrospray ionization mass

1

2

D-glucose D-fructose

2

1

1

2

Deprotonation

transition state

Intermediate-1 Hydride shift

transition state

Reprotonation

transition state

Intermediate-2


represents a metal center [e.g., Cr, Al, and Sn (in Sn BEA)]. Redrawn from Choudhary

et al.14

38

spectrometry (ESI-MS/MS) to propose, due to its abundance, [Al(H2O)4(OH)2]1+ is the

active species when AlCl3 is used as a catalyst.15,128 Given the difference in the active

species between CrCl3 and AlCl3, the fact that ESI-MS/MS may alter speciation

during the measurement,16 and that observed species are often spectators rather than

the active ones, the active aluminum species still remain(s) unknown.

In this work, we integrate kinetics with more direct speciation measurements

for the first time to provide evidence for the active species of AlCl3. Our judicious

choice of AlCl3 is based on the fact that it is a selective catalyst in glucose

isomerization and an excellent salt to be followed spectroscopically.129–133 We

investigate the interplay between Lewis (AlCl3) and Brønsted (HCl) acids using the

OLI software, while simultaneously developing an experimental protocol to quantify

the various aluminum species. We use 27Al quantitative nuclear magnetic resonance

(qNMR) and pH to measure the aqueous aluminum species and protons, respectively,

and combine inductively coupled plasma – mass spectrometry (ICP-MS), dynamic

light scattering (DLS), and ultrafiltration to measure the aluminum nanoparticles.

Finally, we correlate the glucose isomerization rate with the aluminum species

concentrations to propose an active species.

3.2 Methods

3.2.1 Speciation Model

AlCl3 speciation was predicted using the mixed-solvent electrolyte OLI

software.134,135 The model predicts speciation by solving chemical equilibria of

multicomponent systems and combining standard-state thermochemical properties of

solution species with an expression for the excess Gibbs free energy. The standard-

39

state properties are calculated using the Helgeson-Kirkham-Flowers (HKF) equation

of state for aqueous ions and electrolytes at infinite dilution, while the excess Gibbs

free energy model incorporates short, middle, and long-range electrostatic interactions.

Further details regarding these calculations can be found in our previous work.7 The

OLI model has been parameterized at conditions that differ from those typically

encountered in isomerization chemistry and thus, its accuracy is unknown. We address

this point in this paper by more direct measurements of Al speciation.

3.2.2 pH Measurements

pH measurements were obtained using the 3300 High Temperature (HT)

PERpH-X sensor (Rosemount) and the 56 Advanced Dual-Input analyzer

(Rosemount). A typical measurement was conducted in a 100 mL thick-walled glass

vessel (Fisher Scientific) containing catalyst solution. The catalyst solution was stirred

and heated for a specified time. The pH sensor was then inserted into the vessel and

pH was obtained in situ. Heated samples were then cooled to room temperature and

pH was obtained ex situ. The sensor was confirmed to measure the pH of buffer

solutions (Fisher Scientific) within their specified temperature range (298 to 373 K,

see Figure B-1).

3.2.3 27Al Quantitative Nuclear Magnetic Resonance (qNMR) Spectroscopy

27Al qNMR was carried out on an Avance III 400 MHz NMR spectrometer

(Bruker). Spectra were measured at 104.27 MHz (pulse width: 12.50 µs, acquisition

time: 0.1966 s, scans: 64, solvent: 90% H2O, 10% D2O) and processed using

Mestrelab Research software (mNOVA). The samples were prepared in quartz NMR

tubes (NewEra), containing 0.5 mL of reaction mixture and were studied at 303 or 363

40

K. For quantification purposes, an external standard of Al(H2O)63+ was prepared with

5 mM AlCl3 and 100 mM HCl (Figure B-2).

3.2.4 Ultrafiltration

Ultrafiltration was performed using Vivaspin 500 concentrators (Sartorius)

with a 10,000 molecular weight cut-off (MWCO). The concentrators were filled with

500 µL of either freshly prepared or heat-treated catalyst solutions and placed into the

centrifuge (Eppendorf Centrifuge 5424) for 15 minutes at a spin speed of 15,000 g.

Samples were then recovered from the bottom of the concentrate pocket with a pipette.

3.2.5 Inductively Coupled Plasma Mass Spectrometry (ICP-MS)

ICP-MS measurements on freshly prepared and heat-treated catalyst solutions

were carried out using an Agilent 7500cx Series instrument (Wilmington, DE).

Samples underwent ultrafiltration, followed by acid treatment prior to ICP-MS

analysis. The aluminum standards (Agilent) were prepared in HNO3 (70 w/w %,

Sigma Aldrich) and diluted for the ICP-MS calibration.

3.2.6 Dynamic Light Scattering (DLS)

DLS experiments were conducted on a Brookhaven ZETAPALS instrument.

DLS measurements were obtained both before and after heating the catalyst solutions.

The autocorrelation functions were analyzed using the Multimodal Size Distribution

algorithm of the accompanying software, where the viscosity of the solutions was

assumed to equal that of water.

41

3.2.7 Catalytic Measurements

All chemicals were purchased from Sigma Aldrich and used as received. The

reactions were conducted in 10 mL glass vials (Sigma Aldrich) heated in an aluminum

reactor block with controlled stirring (Fisher Scientific). Catalyst solutions of AlCl3∙6

H2O (Sigma Aldrich) and HCl (37 w/w %, Sigma Aldrich) were stirred and preheated

at reaction temperature for 24 h. Kinetic experiments were then carried out in sealed

reactor vials containing the preheated catalyst solution (2 mL) and glucose. The Al-to-

reactant molar ratio was kept at 9:100, which corresponds to 5 mM AlCl3 and 1 wt %

glucose solution. The HCl concentration varied from 3 to 44 mM when used with

AlCl3 (except for the external standard mentioned above). At different time points, the

vials were removed from the oil bath, quenched in ice to stop the reaction, and filtered

with a 0.2 μm filter (Fisher Scientific).

Quantification of the liquid products was achieved using high-performance

liquid chromatography (HPLC, Waters Alliance Instruments e2695), equipped with a

refractive index (RI) detector and a photodiode array (PDA) detector. Sugars were

separated using a Biorad HPX87C column at 348 K with HPLC-grade water flowing

at 0.5 mL min-1 as the mobile phase. Acid byproducts and HMF were separated using

a Biorad HPX87H column heated to 323 K with 5 mM sulfuric acid flowing at 0.5 mL

min-1 as the mobile phase.

3.3 Results and Discussion

3.3.1 Model-Predicted AlCl3 Speciation in Aqueous Media

In aqueous media, AlCl3 dissociates to form metal cations. These cations are

solvated by water, forming complexes, such as [Al(H2O)6]3+, which can be further

hydrolyzed, as shown in Eq. 3-1.126,127,136

42

[Al(H2O)6]3+ + x H2O ↔ [Al(H2O)6-x(OH)x]

(3-x)+ + H3O+ Eq. 3-1

The main species predicted using the OLI software vs. HCl concentration

include the hexa-aqua species, [Al(H2O)6]3+, the stable, crystalline solid form of

aluminum hydroxide, called boehmite, AlO(OH), and the mono- and di-hydroxy

species, [Al(H2O)5(OH)]2+ and [Al(H2O)4(OH)2]1+, respectively (Figure 3-2 and

Figure 3-3). The cations are represented as Al3+, AlOH2+, and Al(OH)21+ (for

simplicity).

With increasing HCl concentration, the Al3+ concentration increases

monotonically, and that of AlO(OH) decreases monotonically, whereas the

concentrations of Al(OH)2+ and Al(OH)21+ exhibit volcano-like curves with peaks at

intermediate HCl concentrations. At low HCl concentrations, the concentrations of

Al(OH)2+ and Al(OH)21+ increase as the pH drops at the expense of the solid. The

distribution of aluminum species at high HCl concentrations can be rationalized by Le

Chatelier’s principle, where an increase in H+ concentration shifts the thermodynamic

equilibrium to the left in Eq. 3-1, suppressing the formation of the hydrolyzed

aluminum species and increasing the concentration of Al3+. The dominance of Al3+

species at high acid concentrations has been reported before.14,126 In addition, the

prediction of solid boehmite is consistent with Hem et al.’s findings, which

demonstrated that heating aqueous aluminum results in the formation of boehmite,

whose structure was confirmed by x-ray diffraction (XRD).137 To our knowledge, the

Mono-hydroxy

(b)

Di-hydroxy

(c)

Hexa-aqua

(a)

Figure 3-2: Al (III) ions generated from the dissolution of AlCl3 in aqueous media.

43

concentrations of the hydrolyzed aluminum species have not been quantified;

however, Mesmer et al. determined the hydrolyzed species are most significant at

relatively low concentrations of aluminum (e.g., ≤10 mM),138 as considered here. The

model predicts the hydrolyzed species’ concentrations increase with an increase in

solution temperature, which is expected given that aluminum hydrolysis is

endothermic (Figure B-3).136 Qualitatively, the results remain similar when

temperature changes.

3.3.2 Equilibration of AlCl3-HCl Catalyst Solutions

Next, we investigated the time necessary for the AlCl3-HCl catalyst solutions

to reach equilibrium. Equilibration allows formation of the various aluminum

species139 and thus proper comparison between the experimentally measured and

predicted speciation,135 and ensures quasi-equilibrated speciation during short reaction

times to enable kinetic analysis (which otherwise can be influenced by varying


(413 K), calculated using the OLI software.

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0

1

2

3

4

5

6

0 20 40 60 80 100

Al(

OH

)2+

(m

M),

Al(

OH

) 21

+(m

M)

Al3

+ (

mM

), A

lO(O

H)

(mM

)

HCl (mM)

Al+3

AlO(OH) - Sol

AlOH+2

Al(OH)2+1

Al3+

AlO(OH)

Al(OH)2+

Al(OH)21+

44

speciation during kinetics measurements). Different approaches have been employed

in prior work to determine the time necessary for equilibrium. Frink et al. measured

the pH of aluminum salt solutions as a function of time and temperature (298 and 313

K) upon cooling (ex situ).139 Hem performed measurements of the pH together with

quantification of polymeric and solid aluminum.137 Since OLI predictions indicate the

formation of solid aluminum, we measure pH and solids. The catalyst solutions were

preheated at reaction temperature for 24 h, and the pH was obtained in situ and ex situ;

not a noticeable difference between in situ and ex situ pH measurements was found

(see Figure 3-4 at 363 K and Figure B-4 for measurements at 303 K and 413 K). The

irreversible pH behavior observed upon cooling is most likely due to the formation of

stable solid aluminum species. Thereafter, we present only ex situ pH data. The solid

was quantified through ultrafiltration followed by ICP-MS analysis of the permeate.

The difference between the initial amount of aluminum, prior to preheating, and the

concentration of aluminum in the permeate was taken as the amount of solid.

0

0.5

1

1.5

2

2.5

3

3.5

4

0 10 20 30 40 50

pH

HCl (mM)

24 hr

Series3

Series1

Ex situ (pre-heat: 24 h, 363 K)

In situ, 363 K

Speciation model, 363 K

Figure 3-4: pH measured ex situ (closed circles) or in situ (open circles) compared to

pH calculated from OLI software (line) at 363 K.

45

Figure 3-5a shows the ex situ measured pH and Figure 3-5b shows the amount

of solid as a function of time. The pH initially drops during the first 1 h and then

remains approximately unchanged after 8 h, whereas solid formation increases with

time and eventually remains constant after 24 h. Figure 3-5 indicates multiple time

scales whereby solids form slowly compared to equilibration of small species.

Hsu proposed the solid results from an intermediate polymeric species,125

which forms within the first hour of heating, and, in combination with the formation of

the hydrolyzed species, results in the formation of H+, which decreases the pH. As

time increases, the polymeric species undergo nucleation to form stable solid species,

such as boehmite.125 This nucleation and/or subsequent growth is slow, given the

number of particles continue to increase for 24 h. DLS measurements further indicate

the formation of nanoparticles and provide the average particle growth as a function of

time (Figure B-5). The average particle diameter remains constant after 24 h of

heating, which is consistent with the time taken for observable solid formation using

Figure 3-5: Determination of equilibrium from (a) pH, measured ex situ, and (b)

amount of solid formed with time. Samples contained 5 mM AlCl3 and were cooled

from 413 K to 303 K.

1

2

3

4

5

0 8 16 24 32 40 48

pH

Time (hr)

0 mM HCl

3 mM HCl

10 mM HCl

20 mM HCl

44 mM HCl

0

1

2

3

4

0 8 16 24 32 40 48

So

lid

Al

(mM

)

Time (hr)

0 mM HCl

3 mM HCl

10 mM HCl

20 mM HCl

44 mM HCl

(a) (b)

46

ultrafiltration/ICP-MS. We observed similar results for samples preheated at 363 K

(Figure B-5). For these samples, solids were present after being heated for 24 h or

longer, and their size and amount remained constant thereafter whereas the pH

equilibrated faster (Figure B-6). Therefore, we have decided to preheat our samples

for 24 h at reaction temperature for equilibration of aluminum species prior to

conducting kinetic experiments.

3.3.3 Direct Measurements of AlCl3 Speciation

Figure 3-6 outlines the methodology to quantify each species. qNMR

measurements were performed in situ at 363 K and ex situ at 413 K due to instrument

limitations.

Figure 3-6: Methodology to quantify aluminum speciation at specified concentrations

of AlCl3 and HCl, temperature, and heating time.

47

3.3.3.1 Hexa-Aqua Aluminum (Al3+)

We first quantified the hexa-aqua monomer, Al3+, using 27Al qNMR, at a

chemical shift of ~0 ppm.140 qNMR spectroscopy is non-destructive, quantitative, and

uses a low-energy electromagnetic wave, causing an extremely small energy

perturbation to the system.72,141,142 In addition, the quantitative accuracy of NMR

analysis has reached three significant figures.72,141,142 Following a similar approach to

that of Maki et al.,72 we measured the concentration of Al3+ by comparing the qNMR

peak areas of experimental samples to an external standard, which consisted of 5 mM

AlCl3 and 100 mM HCl. The concentration of HCl in the standard is high to ensure all

aluminum is in the form of Al3+, according to Figure 3-3. Similar to the pH

measurements, it did not make a difference if the Al3+ species was quantified in situ

(Figure B-7, Table B-1) or ex situ (Figure 3-7, Table B-2); both measurements were in

reasonable agreement with the speciation model (Figure 3-9a). We hypothesize the

Figure 3-7: 27Al qNMR spectra for Al3+. Experimental conditions: preheat 413 K for

24 h, 5 mM AlCl3. qNMR measurements obtained upon cooling to 303 K.

δ (ppm)

(f) 100 mM HCl

external standard

(e) 44 mM HCl

(d) 20 mM HCl

(c) 10 mM HCl

(b) 3 mM HCl

(a) 0 mM HCl

48

unaccounted aluminum exists primarily as suspended solids, which are not detect by

liquid-phase qNMR.126

3.3.3.2 Solid Aluminum

We next quantified the solid aluminum species. Techniques such as dialysis

and ultrafiltration are commonly used to separate fine colloidal mineral aluminum and

polymeric aluminum species from soluble aluminum.126 Therefore, we combined

ultrafiltration with DLS and ICP-MS measurements to quantify solid aluminum

species in our catalyst solutions. DLS measurements detected the presence of

aluminum particles and provided the average particle diameter. This allowed for

selection of an ultrafiltration membrane with a pore diameter of 5 nm, significantly

less than that detected by DLS (Figure B-8). The samples then underwent heating and

ultrafiltration. ICP-MS determined the concentration of aluminum in the permeate.

Our experimentally inferred data are consistent with the calculated amount of solid

aluminum (Figure 3-9b).

3.3.3.3 Hydrolyzed Aluminum Monomers (AlOH2+ and Al(OH)21+)

The low concentration and rapid proton exchange associated with the

hydrolyzed species make them difficult to observe experimentally.136 Therefore, we

deduced their concentration through the fundamental equilibrium relation (Table B-3

and ). This required the Al3+ and H+ concentrations and the acid dissociation constants

(KA,1 and KA,2). The Al3+ and H+ concentrations were obtained by qNMR and pH

measurements, respectively. The acid dissociation constants were obtained from the

OLI software (Table B-3 and Table B-4). The experimentally estimated mono- and di-

hydrolyzed concentrations exhibit the same volcano-like behavior predicted by the

49

OLI software as a function of HCl concentration, as shown in Figure 3-9c and Figure

3-9d, respectively.

3.3.3.4 Glucose Conversion and Catalytically Active Species in Equilibrated

Catalyst Solution

A series of kinetic experiments with glucose added to preheated catalyst

solutions containing AlCl3 and HCl were performed. The initial rates of glucose

isomerization were obtained by fitting the glucose concentration profiles at low

conversions (≤15%) with a first-order rate expression, where fructose was the main

product observed (Figure B-9 and Figure B-10). Figure 3-8 shows these rates as a

function of HCl concentration.

Interestingly, with increasing HCl, the glucose conversion rate exhibits a trend

that is qualitatively consistent with the Al(OH)2+ and Al(OH)21+ species’

concentrations, shown in Figure 3-3 (see Figure B-10 for 363 K data). The slow

glucose consumption at low HCl concentrations (before the maximum rate) is

Figure 3-8: Effects of HCl on glucose isomerization rate in AlCl3 solutions (5 mM)

that have been equilibrated at (a) 413 K and (b) 363 K for 24 h prior to kinetic study.

Reaction conditions: glucose 1 wt %, Al to glucose 9 : 100.

0.001

0.003

0.005

0.007

0.009

0 10 20 30 40 50

r glu

cose

(min

-1)

HCl (mM)

413 K

0.0001

0.0002

0.0003

0.0004

0 10 20 30 40 50

r glu

cose

(min

-1)

HCl (mM)

363 K

(a) (b)

50

congruent with the formation of solid aluminum species, indicating the solid likely

hinders glucose conversion. Tang et al. observed, when using solid Al(OH)3 as a

catalyst, that Al(OH)3 did not effectively catalyze the reaction, and resulted in low

glucose conversion (X = 7.8%) and fructose yield (YFru = 6.5%) compared to AlCl3 (X

= 31.8% and YFru = 26.3%) at the same reaction conditions.15 Similarly, the Al3+

species is not likely catalytically active given the reaction rate continues to decrease as

the Al3+ species concentration increases. The correlation between the estimated

glucose consumption rate and the hydrolyzed species strongly indicates that one or

both of these Al(OH)x(3-x)+ species may be catalytically active for glucose

isomerization. Since the speciation model135 was developed at conditions different

from typical sugar experiments, we used our direct speciation measurements to

elucidate the aluminum species. At 413 K, the linear scaling observed between the

glucose isomerization rate and the di-hydrolyzed Al(OH)21+ concentration (Figure 3-9)

strongly indicates Al(OH)21+ is the catalytically active species.

51

At 363 K, the low reaction rate, extremely small values of the rate constants

and low concentrations of the aluminum species (see Table B-4 and Figure B-11)

render our data inconclusive regarding the active species.

0

0.002

0.004

0.006

0.008

0 0.25 0.5 0.75 1

r glu

cose

(min

-1)

Al3+ / Al3+(max)

0

0.002

0.004

0.006

0.008

0 0.25 0.5 0.75 1r g

luco

se(m

in-1

)Solid Al / Solid Al (max)

0

0.002

0.004

0.006

0.008

0 0.25 0.5 0.75 1

r glu

cose

(min

-1)

Al(OH)2+ / Al(OH)2+(max)

0

0.002

0.004

0.006

0.008

0 0.25 0.5 0.75 1

r glu

cose

(min

-1)

Al(OH)21+ / Al(OH)2

1+(max)

(a) (b)

(c) (d)

Figure 3-9: Glucose conversion as a function of measured Al species’ concentrations,

normalized to the maximum observed species’ concentration (see Table B-3 for

observed species’ concentrations). Catalyst solutions were preheated at 413 K for 24 h

prior to kinetic study. Reaction conditions: glucose 1 wt %, Al to glucose molar ratio

of 9 : 100, 413 K.

52

3.4 Conclusions

We have coupled kinetic studies with direct speciation measurements to

elucidate the active species of AlCl3 in glucose to fructose isomerization. Experiments

were designed based on insights obtained from modeling the aluminum hydrolysis

using the OLI software. We established an experimental protocol to quantify the

various aluminum species that employs 27Al qNMR for the hexa-aqua aluminum

monomer, ultrafiltration and ICP-MS for the solid, and pH measurements combined

with equilibrium relations for the mono- and di-hydroxy aluminum species. We found

that speciation reaches equilibrium after heating the catalyst solutions for long times.

Linear scaling between the glucose isomerization rate and the speciation

measurements at sufficiently high temperatures indicates that the hydrolyzed Al(III)

complex [Al(H2O)4(OH)2]1+ is the active species in glucose isomerization. This

finding supports the hypothesis of Hu and co-workers, who observed

[Al(H2O)4(OH)2]1+ using ESI-MS/MS.15 Furthermore, our findings support the idea

that the catalytic species responsible for the isomerization exists as a bifunctional

Lewis acidic/Brønsted base site.14 The approach developed here can be applied to

study the speciation of other metal halides. For metals not observed by qNMR, such as

CrCl3, other spectroscopic techniques, such as ultraviolet visible light spectrometry

(UV-Vis), could possibly be employed to study the dominant hexa-aqua species.

53

BRANCHED BIOLUBRICANT PRODUCTION THROUGH ALDOL

CONDENSATION

4.1 Introduction

Innovations in transportation, industrial production, and alternative energies

have led to an increased demand for lubricants. By 2025, the lubricants’ market is

projected to be worth $166.25 billion USD, up by about $47 billion USD compared to

2016.143 Increased global demand, combined with higher consumption index and

economic growth, warrant continued development of lubricants that are high-

performing and result in minimal harmful impacts on the environment.

Base oils play a major role in determining a lubricant’s performance. They

account for the majority of a lubricant’s cost and comprise anywhere between 75 and

99% of a lubricant’s formulation.19 According to the American Petroleum Institute

(API), there are five categories of mineral base oils (Groups I – V).17 Base oils from

Groups I through III are obtained by solvent-refining, distillation, or hydro-processing

of petroleum.18 Base oils from Group IV have undergone chemical upgrading; for

example, 1-decene from petroleum undergoes oligomerization to form poly-α-olefins

(PAOs).19 Finally, Group V includes all other oils.17 ExxonMobil Basestocks 2018

Pulse Report suggests Group III+ oils will experience the greatest increase in demand

over the next 10 years (+4% increase) owing to their high fuel efficiency and quality.20

Nevertheless, their production is expensive and energy intensive, requiring petroleum

feedstocks, which contribute to greenhouse gas emissions, and harsh reaction

Chapter 4

54

conditions, especially when making Group IV PAOs, the synthesis of which requires

corrosive catalysts (AlCl3, BF3, and HF).18,21 To mitigate these challenges, promising

renewable alternatives from bioderived feedstocks have gained momentum.22–29

Biobased feedstocks can result in a nearly “closed carbon balance” through CO2

capture during photosynthesis and have unique functional groups that enable site-

specific chemistries during processing.

In one of the early attempts to make biobased products, Corma et al. produced

n-tricosane, a linear alkane product with 23 carbons, from lauric acid, a fatty acid

found in coconut and palm kernel oils.26 This process involved ketonization of lauric

acid to form 12-tricosanone, containing 23 carbons, followed by its

hydrodeoxygenation (HDO) to produce n-tricosane (58.2% selectivity) and C10

through C22 alkanes (13.9% selectivity total). Although the final product would not be

suitable as a Group III base oil (18 to 40 carbons) owing to its poor viscosity and high

melting point, the short-chain alkanes may be a suitable for ultra-low sulfur diesel

(ULSD) fuel. Even so, USLD blends ($3.15 per gallon)30 are not as price competitive

as Group III base oils ($5.01 per gallon).31 Despite these obstacles, the strategy

proposed by Corma et al. is appealing. The intermediate ketone, 12-tricosanone, can

be obtained in 89% yield by ketonization of lauric acid and is an ideal starting material

for lubricant synthesis because it contains a high number of carbon atoms and can

partake in carbon-carbon coupling reactions due to the presence of a ketonic group.

Ketones are excellent platforms to incorporate branching because of their

acidic -CH group at the α carbon position.28,29,32–35 Previously, Bell and co-workers

performed multiple cross-ketonization reactions, starting from short-chain length C3 –

C5 carboxylic acids, followed by HDO, to produce C12 – C33 branched and cyclic

55

alkane diesel fuels and lubricant base oils.27 Wang and co-workers produced C23

biolubricant base oil with ~50% yield from acetone and furfural via successive aldol

condensation and HDO steps.25 Although these routes use renewable feedstocks, for

example, carboxylic acid and acetone, which can be produced from sugar

fermentation, or lignocellulosic biomass-derived furfural, they require multiple

reaction steps and extractions. This results in carbon loss, the need for excessive

amounts of solvent, and high production costs.

In the present work, we combine strategies to synthesize a highly branched

biolubricant base oil in fewer steps using 12-tricosanone and furfural as the starting

feedstock. The synthesis of 12-tricosanone in high yield from lauric acid is discussed

above.26 We perform selective aldol condensation followed by HDO to afford the final

products, C28 and C33 branched alkanes (Figure 4-1), having comparable viscosity to

petroleum-derived Group III and Group IV base oils.

56

4.2 Methods

4.2.1 Materials

Furfural and 12-tricosanone were purchased from Tokyo Chemical Industry

Co. Methanol (≥99.8%), sodium hydroxide pellets, hydrochloric acid (36.5-38.0%),

and cyclohexane (99.9%) were purchased from Fisher Scientific. Eicosane (99%) and

hydrogen hexachloroiridate hydrate (99.98%, metal basis) were purchased from

Sigma-Aldrich. Ammonium perrhenate(VII) (99.999% metals basis) was purchased

from Alfa Aesar. Fuji Silysia Chemical Ltd. G6 (BET surface area, 535 m2/g) kindly

supplied silica gel.

Figure 4-1: Renewable approach to produce biolubricant base oil. Base-catalyzed aldol

condensation (AC) of furfural and 12-tricosanone to form a C33 furan intermediate

along with a small fraction of C28 intermediate, followed by their HDO over an Ir-

ReOx/SiO2 catalyst to produce base oils containing C33 branched alkane as the major

component.

57

4.2.2 Catalyst Preparation

The Ir-ReOx/SiO2 (Ir 4 wt% loading, Re/Ir =2, molar) catalyst was prepared by

a sequential impregnation method.144 First, Ir/SiO2 was prepared by impregnating Ir on

SiO2 (Fuji Silysia G-6) with an aqueous solution of hydrogen hexachloroiridate

hydrate. Next, the solvent was evaporated at 75 °C on a hotplate (IKA) and dried at

110 °C for 12 h in an oven (Fisher Scientific). The resulting Ir/SiO2 was impregnated

with ReOx by using an aqueous solution of ammonium perrhenate(VII). The catalyst

was calcined in a crucible in air at 500 °C for 3 h at a 10 °C/min temperature ramp.

The reported metal loadings in the catalyst are based on the theoretical amount of

metals used for impregnation. The catalyst was used in powder form with a granule

size of <400 mesh. BET surface area measurements were performed on a

Micromeritics ASAP 2020 Accelerated Surface Area and Porosimetry instrument at

the Advanced Materials Characterization Laboratory at the University of Delaware.

The measured BET surface area of Ir-ReOx/SiO2 was 368.6 m2/g.

4.2.3 Reaction Procedures

4.2.3.1 Aldol Condensation

All reactions were conducted in 10 mL glass vials (Sigma-Aldrich) heated in

an aluminum heating block with controlled stirring (Fisher Scientific). In a standard

reaction, the reactants, furfural (4.68 mmol, 0.45 g) and 12-tricosanone (0.295 mmol,

0.10 g), were combined with the solvent, methanol (5 mL), and catalyst, sodium

hydroxide (1 M, 0.20 g). The catalyst concentration was selected from a prior report.25

Temperature and stirring rate were maintained at 80 °C and 400 rpm, respectively.

After reaction for the set time, the reaction vials were removed from the heating block

and cooled to room temperature. Methanol and unconverted furfural were removed by

58

rotatory evaporation (Buchi). The products were washed and neutralized with a dilute

solution of hydrochloric acid (1 M) and extracted with dichloromethane (20 mL).

Dichloromethane was removed by rotary evaporation prior to HDO. The products

formed by aldol condensation were solids at room temperature.

4.2.3.2 Hydrodeoxygenation (HDO)

HDO reactions of aldol condensation products were conducted in a 50 mL Parr

reactor with an inserted Teflon liner and a magnetic bar. The catalyst, Ir-ReOx/SiO2

(0.15 g), and solvent, cyclohexane (5 mL), were added to the reactor for catalyst pre-

reduction. The reactor was sealed with a fitted reactor head that contained a

thermocouple, a rupture disk, a pressure gauge, and a gas release valve. The mixture

was heated at 200 °C and 5 MPa H2 for 1 h at 240 rpm. Upon pre-reduction, the

reactor was cooled to room temperature, and H2 was released. Next, the aldol

condensation products (0.40 g) were mixed with cyclohexane (15 mL) and added to

the reactor. The reactor head was immediately closed, purged with 1 MPa H2 three

times, and pressurized to 5 MPa H2. The reaction occurred over 18 h at 180 °C and

with continuous stirring at 500 rpm. The heating time to reach the set temperature was

approximately 25 min and was not included in the total reaction time. Upon

completion, the reactor was immediately transferred to a water bath. The catalyst was

separated from the solution by centrifugation.

4.2.3.3 Analysis of Products

The products were analyzed by GC (Agilent 7890A), equipped with an HP-1

column and a flame ionization detector. Products were quantified using eicosane (C20)

as the internal standard (0.10 g). The products were identified by GC-MS (Agilent

59

7890B and Agilent 5977A with a triple-axis detector), equipped with a DB-5 column.

High-resolution MS liquid-injection field desorption ionization (HRMS-LIFDI;

Waters GCT Premier) data were obtained from the Mass Spectrometry Facility at the

University of Delaware. 1H and 13C NMR spectroscopy (Bruker AV400, CDCl3

solvent) was also used to identify the products from aldol condensation and HDO.

The conversion and the yield of all products from aldol condensation and HDO

reactions were calculated on a carbon basis using the following equations:

Conversion [%-C]=ninitial reactant - nunreacted reactant

ninitial reactant

× 100 Eq. 4-1

Yielddetected products [%-C]=nproduct × c atoms in product

ntotal C atoms in 12-tricosanone × 100 Eq. 4-2

4.2.4 Lubricant Properties

Cyclohexane and lower boiling point alkanes (<C6) in the products were

removed by rotary evaporation prior to viscosity measurements. The lubricant

properties of the final HDO products were evaluated according to the American

Society for Testing and Materials (ASTM) methods. The kinematic viscosities at 100

and 40 °C (KV100 and KV40) were determined following the ASTM D445 method.

The viscosity index (VI) was calculated using the KV100 and KV40, following the

ASTM D2270 method. An extra-low-charge-semi-micro viscometer (Cannon, size

150, calibrated model #: 9722-H62) apparatus was used for all measurements (Figure

C-1). The sample charge volume was 300 µL. To demonstrate the accuracy of our

method, we measured the viscosity of a Cannon N35 Standard and an Exxon Mobil

SpectraSyn PlusTM PAO-4 (Group IV) and found that our measurements were in

agreement with the reported values (Table C-1).

60


4.3.1 Reaction Conditions for Aldol Condensation

We began with the aldol condensation reaction between furfural and 12-

tricosanone. All reactions were catalyzed by 1 M NaOH (unless otherwise specified),

and the reaction temperature was 80 °C for all experiments to ensure 12-tricosanone

(melting point = 68 °C) was a liquid at reaction temperature. Our goal was to identify

a suitable solvent and the best reaction conditions (reaction time, feed ratio) to obtain

high yields of the desired C28 and C33 furan intermediates, especially the C33

intermediate, which has two branches.

Figure 4-2 shows the yield of products and conversion of 12-tricosanone in

different solvents. Previous works have shown that aldol condensation may occur

with25 or without a solvent.145 We found low conversion of 12-tricosanone (<25%)

and no detected C28 or C33 furan intermediates without a solvent, indicating the

observed conversion of 12-tricosanone was likely due to its oligomerization with itself

and/or furfural. Del Rio and Philip suggested that carboxyl and hydroxyl groups may

promote oligomerization to form high-molecular-weight-organic compounds found in

oils and source rocks.146 These oligomers, however, are difficult to detect by

conventional GC columns owing to their high molecular weights (>C40).144,147

Similarly, we believe the carboxyl and hydroxyl groups in 12-tricosanone and furfural

promote the formation of high molecular weight, solid oligomeric species, called

humins, whose characterization is reported in our prior work.8

Because the absence of solvent led to low conversion of 12-tricosanone, we

next screened different solvents as a function of their polarity. Non-polar, aprotic

solvents, cyclohexane and dioxane, resulted in very low yield of condensation product,

61

which is likely due to their poor ability to dissolve the reactants and catalyst,

combined with their inability to donate protons. In contrast, a polar, protic solvent,

methanol, dissolved the reactants and catalyst and resulted in the highest conversion of

12-tricosanone. Given the propensity of methanol, we also tested water, but it resulted

in no yield of product. The observed conversion of 12-tricosanone with water as the

solvent is likely owing to the formation of humins. Water was most likely ineffective

because it is a product of aldol condensation, and when used as a solvent, it shifts

equilibrium towards the reactants, according to Le Chatelier’s principle. Therefore, we

performed subsequent aldol condensations in methanol.

0

20

40

60

80

100

Yie

ld &

Con

ver

sion

of

12-t

rico

san

on

e (%

)

Solvent

C-33 Furan C-28 Furan ConversionYield of C33 furan Yield of C28 furan Conversion of 12-tricosanone

Figure 4-2: Conversion of 12-tricosanone as a function of solvent. Reaction

conditions: furfural (0.227 g) and 12-tricosanone (0.1 g) (mole ratio = 8 : 1), 80 °C, 24

h. Studies performed with no co-solvent, cyclohexane, and dioxane contained 1 M

NaOH (100 µL). Methanol and water studies contained 1 M NaOH (5 mL).

62

Next, we conducted experiments to determine an optimal reaction time and

feed ratio to achieve high yields in desired condensation product, that is, the C33 furan

intermediate. Initially, reactions were performed for 24 h and with mole ratios of

furfural:12-tricosanone ranging from 1:1 to 16:1 (Figure 4-3). GC-MS detected two

fragments with m/z = 416 and nine fragments with an m/z = 464, which correspond to

geometric isomers of the C28 and C33 furans, respectively (Figure C-3). We considered

all isomers corresponding to the same m/z fragment as one product for yield

calculation. Additional analysis by HRMS-LIFDI (Figure C-4) and 1H and 13C NMR

were consistent with the C28 and C33 products identified from GC-MS (Figure C-5 and

Figure C-6).

0

20

40

60

80

100

1:1 2:1 6:1 8:1 16:1

Yie

ld &

Co

nv

ersi

on

of

12

-tri

cosa

no

ne

(%)

Mole Ratio Furfural:12-Tricosanone

C-33 Furan C-28 Furan Total ConversionYield of C33 furan Yield of C28 furan Conversion of 12-tricosanone

Figure 4-3: Yields of C28 and C33 furan intermediates from the reaction of 12-

tricosanone (0.10 g) with furfural at 80 °C and 1 M NaOH in methanol for 24 h.

63

Importantly, the product distribution could be tuned depending upon the feed

ratio, for which high amounts of furfural favored the C33 furan (YC33furan = 64.8% for

furfural/12-tricosanone=8:1, Figure 4-3). The yields of total detected C28 and C33

furans are lower than the 12-tricosanone conversion, which could be due to the

formation of humins. Higher selectivity to desired condensation products was

achieved by decreasing the reaction time from 24 h (Figure 4-3) to 8 h (Figure 4-4). A

maximum 94.3% yield of branched furan intermediates containing C33 as the major

product (YC28furan = 14.8% and YC33furan = 79.5%, Figure 4-4) was obtained for 8 h and

at furfural/12-tricosanone molar ratio of 16:1. Excess furfural can be recycled upon

64

separation by rotary evaporation. A higher feed ratio beyond 16:1 was detrimental to

the yield of C33 furan and the overall yield (Figure 4-4).

4.3.2 HDO to Produce Branched Alkane Base Oil

The aldol condensation product, after neutralizing and removing methanol and

any unconverted furfural, was dissolved in cyclohexane and transferred to the Parr

reactor for HDO. Our reported best reaction conditions to produce jet fuels144 and

lubricants22 with Ir-ReOx/SiO2 as the catalyst were adapted. Liu et al. determined that

Ir and ReOx act synergistically, during which Ir initiates the hydrogenation steps by H2

dissociation, and the acid sites of partially reduced ReOx activate etheric C-O bonds in

0

20

40

60

80

100

1:1 2:1 6:1 8:1 16:1 25:1

Yie

ld &

Co

nv

ersi

on

12

-tri

cosa

no

ne

(%)

Mole Ratio Furfural:12-Tricosanone

C33 furan C28 furan Total ConversionYield of C33 furan Yield of C28 furan Conversion of 12-tricosanone

Figure 4-4: Yields of furan intermediates from the reaction of 12-tricosanone (0.10 g)

with varying amounts of furfural at 80 °C and with 1 M NaOH in methanol (5 mL) for

8 h. Error bars correspond to mean ± standard error of the mean (SEM) of three

independent reactions.

65

the furan ring.144 Our prior X-ray photoelectron spectroscopy (XPS) characterization

of Ir-ReOx/SiO2 showed that Re undergoes reduction to form Re2+ and Re4+ (41% and

11%, respectively) and Re0.144 In addition, pyridine adsorption studies, monitored by

Fourier-transform infrared spectroscopy (FTIR), were performed in our prior work to

distinguish acid sites on the in situ-reduced catalyst.144 The FTIR data showed that the

catalyst contains mainly Lewis acid sites (1490 cm-1), which are likely the partially

reduced ReOx species, and a weak but distinctive band at 1540 cm-1 for the Brønsted

acid sites.144

HDO of C28 and C33 furans mixture (YC28furan = 14.8% and YC33furan = 79.5%)

over Ir-ReOx/SiO2 yielded 61.4% of total lubricant-ranged branched alkanes (C28 and

C33) at 180 °C and 5 MPa H2 in cyclohexane for 18 h. The major products are C33

(YC33 = 50.5%) and C28 (YC28

= 10.9%) alkanes, followed by C15, C14, C13 and C10

alkanes (𝑌combine < 11%), which likely result from carbon-carbon (C-C) cracking in the

tertiary and secondary carbon atoms of the C33 and C28 alkanes (Figure C-2, Figure C-

3). Similar C-C bond breaking was observed in prior work for HDO of a trifuran.144 In

this work, C5 alkane is also likely to form through C-C cracking (Figure C-2), but it

was not quantified because of difficulty arising from the solvent, cyclohexane,

forming C1 – C6 alkanes during pre-reduction of the catalyst and during the HDO

reaction, as demonstrated in our prior work.144 Thus, alkanes with less than six carbon

atoms were not quantified. In addition to C5 alkane, gaseous C1 – C4 alkanes could

also form from the substrates in the HDO step to account for some of the carbon loss,

but gas-phase products were not quantified in this work. Furthermore, coke formation

on the catalyst and unidentified oxygenated intermediates may contribute to the

remaining carbon loss. Coke formation is not expected to hinder recycling the catalyst

66

because Liu et al. have demonstrated that upon recalcination the catalyst can be

regenerated with similar activity to that of the fresh catalyst.144 GC-MS and HRMS-

LIFDI detected mass fragments expected for C28 and C33 alkane products (Figure C-3

and Figure C-4), and the NMR spectra were congruent with that expected for the C33

alkane product (Figure C-7 and Figure C-8).

4.3.3 Lubricant Properties

The biolubricant alkane base oil obtained by our method is a viscous liquid at

room temperature and even at low temperature of approximately 3 °C, whereas n-

tricosane, containing 23 carbons and produced by HDO of 12-tricosanone, is solid.26

This suggests the addition of branches on linear alkanes by our method can

significantly improve the low-temperature flow property. The viscous properties of

our synthesized base oil are determined by extra-low charge, micro-viscosity

measurements and are compared to those of Phillips 66 Ultra-S 3TM (Group III) and

Exxon Mobil SpectraSyn PlusTM PAO-3.6 (Group IV) base oils. When considering

lubricant base oils, at high temperatures (100 °C), an oil’s viscosity (KV100) should

be high to create a thick hydrodynamic film between surfaces. Upon decreasing

temperature (40 °C), an oil’s viscosity (KV40) should be low, or less resistant to flow,

to promote fluidity. The viscosity index (VI), calculated using the KV100 and KV40

values, measures viscosity stability of a fluid as a function of temperature.

Maximizing VI ensures that the oil’s viscosity varies as little as possible with

temperature. Table 4-1 shows that KV100, KV40, and VI of our base oil are

comparable to commercial Group III and Group IV base oils. The VI of our base oil is

115, which is higher than 100, suggesting our oil has high quality.21

67

Table 4-1: Properties of alkane base oil (YC33 = 50.5% and YC28

= 10.9%) compared to

Group III and Group IV commercial lubricants.

[a] KV40 and KV100 are kinematic viscosities at 40 °C and 100 °C, respectively

(ASTM D445). [b] The properties of commercial products were obtained from the

product specifications datasheet disclosed by the manufacturers. [c] The properties of

commercial products were obtained from the product specifications datasheet

disclosed by the manufacturers.

4.4 Conclusions

We demonstrated a viable route to obtain a branched biolubricant base oil in

fewer steps from a long‐chain ketone, 12‐tricosanone, which can be obtained from

fatty acid, and furfural, which can be obtained from lignocellulosic biomass. This

approach involved aldol condensation followed by hydrodeoxygenation (HDO). Aldol

condensation of 12‐tricosanone and furfural at 16:1 molar ratio resulted in a maximum

93.5% yield of branched C28 and C33 furan intermediates in 8 h. The C33 furan

intermediate was the major product (79.5%), and the remaining product was the C28

furan intermediate. Subsequent HDO of aldol condensation product over an Ir‐

ReOx/SiO2 catalyst (Re/Ir molar ratio=2) yielded 72% total alkanes, in which

lubricant‐ranged branched alkanes, C28 and C33, were obtained in 61% yield. Small

amounts (11%) of low‐carbon alkanes (>C10) were detected. Micro‐viscosity

measurements indicated our base oil has viscous properties comparable to petroleum-

derived Group III and IV oils. The strategy to synthesize renewable base oils

Base Oil KV100[a]

[mm2 s-1] KV40[a]

[mm2 s-1]

VI[b]

Our Base Oil 3.43 14.37 115

Phillips 66 Ultra-S 3TM (Group

III)[c] 3.26 13.20 116

Exxon Mobil SpectraSyn

PlusTM PAO-3.6 (Group IV)[c] 3.6 15.4 120

68

described in this paper could be a potential stepping‐stone to replace petroleum‐

derived base oils, which in turn can reduce greenhouse gas emissions.

69

REVERSIBLE FORMATION OF SILANOL GROUPS IN TWO-

DIMENSIONAL SILICEOUS NANOMATERIALS UNDER MILD

HYDROTHERMAL CONDITIONS

Reproduced with permission from the Journal of Physical Chemistry C, in press.

Unpublished work copyright 2020 American Chemical Society.

5.1 Introduction

Silica (SiO2) is the most abundant oxide in the earth’s crust. It is used in many

applications, including as catalysts,148,149 supports,50 separation membranes,48 and

sorbents.150 Upon repeated use in some of these applications, siliceous materials

undergo regeneration to remove carbon-containing deposits from their pores.151

Hydrothermal treatment is an example of regeneration, involving steaming materials

for an extended time at high temperatures. Previous solid-state NMR experiments

have shown that well-ordered siliceous materials lose their crystallinity after being

exposed to steam for extended times, e.g., for 80 days at 823 K and 11 bar of steam.36

Recent work has shown that, even though changes related to steaming may not be

evident using traditional characterization methods, there could be important structural

changes affecting catalyst activity.37 Calcination is another method of regeneration, in

which carbonaceous byproducts react with air at elevated temperatures to form carbon

dioxide and trace amounts of steam. Steam at low concentrations is likely to interact

with silica’s surface; however, these atomic-level interactions are challenging to detect

by methods used to study powders, such as solid-state NMR and XRD, which are

Chapter 5

70

better suited to study the material’s bulk rather than its surface.38,39 These surface

interactions are important, as numerous technological applications of silica rely on its

specific surface properties. In particular, surface silanol groups (SiOH) serve as

anchoring points for a variety of chemical species.40

Surface science techniques, such as Infrared Reflection Absorption

Spectroscopy (IRRAS) and X-ray Photoelectron Spectroscopy (XPS), offer

extraordinary precision but work best when the surface is electrically conductive.

Unfortunately, siliceous materials are not electrically conductive, and many of those

techniques are not applicable to these materials.41 One way to overcome this lack of

conductivity is by casting or growing 2-D silicate thin films onto metal single crystal

substrates. An example of this is the bilayer silicate, which can be prepared on various

metal supports, but Ru(0001) has received the most attention.42–45 The bilayer silicate

consists of corner-sharing tetrahedral [SiO4] building blocks arranged in a honeycomb

structure (Figure 5-1) and can contain amorphous regions in addition to the crystalline

bilayer.44 By “amorphous,” we refer to small clusters above the bilayer, not the 2-D

vitreous regions commonly found within the bilayer structure. Polymorphous areas are

also found in metals supported on siliceous supports.46 The bilayer does not, however,

contain the wide range of 3-D pores and channels present in siliceous zeolites.

Recently, all-Si MFI nanosheets have been deposited onto metal substrates for surface

science studies, especially onto Au(111), which has the benefit of being inert in

nature.47–49 MFI nanosheets contain interconnected pore networks (Figure 5-1) and

have the same framework as ZSM-5, one of the most important zeolites for catalysis

applications.152 The deposition of 2-D siliceous materials onto conductive supports

now provides a model system that can be investigated through the lens of surface

71

science. While many surface science techniques have been traditionally limited to

ultra-high vacuum (UHV) conditions, technical developments in the last few decades

have allowed operation at temperatures and pressures more relevant to practical

applications.52

In this work, we use IRRAS to study the effects of mild hydrothermal

conditions, in situ, on the MFI nanosheets supported on Au(111) and a polymorphous

bilayer silicate supported on Ru(0001). We select these two materials because they

represent model systems for 3-D ones. That is, the polymorphous bilayer silicate can

be analogous to some clays or amorphous siliceous materials used as supports for

metals in catalysis, while the MFI nanosheets are similar to well-ordered, porous

(a) (b)

Sid

e v

iew

To

p v

iew

(a)

(b) (c)

c a

b

7 nm (3.5 unit cell)

c a

b

Figure 5-1: Side views of the MFI nanosheets with (a) 7 nm thickness, which

corresponds to 3.5 unit cells and is the visual representation of the nanosheets used in

this work and (b) a 1.5 unit cell, provided to better see the framework in (a) and which

corresponds to the dashed box in (a). (c) Side view of the bilayer silicate. Red, yellow,

and white atoms correspond to oxygen, silicon, and hydrogen, respectively. Terminal

SiOH groups are present in the MFI nanosheets, but not in the bilayer silicate.

72

siliceous zeolite supports. We find that at 473 and 573 K and 3 mbar H2O, the number

of silanol groups (SiOH) increases in the MFI nanosheets, but does not change for the

case of polymorphous bilayer silicate. The effects of mild hydrothermal conditions in

the MFI nanosheets are reversible and do not result in framework degradation. The

work shown here improves the fundamental understanding of the effects of mild

hydrothermal conditions on 2-D siliceous nanomaterials and serves as a starting point

when considering these effects on 3-D ones.

5.2 Methods

5.2.1 MFI Nanosheets

Pure silica (no Al) zeolite nanosheets with structure-type MFI (MFI nanosheet)

were prepared by seeded growth with an organic structure-directing agent (OSDA,

bis-1,5(tripropyl ammonium) pentamethylene diiodide, denoted as dC5), as reported

previously;49 however, the synthetic conditions were slightly modified to improve

thickness uniformity. Briefly, the MFI nanosheets were fractured using a horn-

sonicator and separated from the large aggregates or MFI nanocrystals by

centrifugation. The nanosheet fragments were then processed with dC5-silica sol at

428 K for 3 days, yielding rectangular nanosheets with ~2 µm lateral dimension and 7

nm predominant thickness. Using the floating particle coating method,153 the

crystallized MFI nanosheets were transferred to a Au(111) crystal, which had been

cleaned through repeated cycles of Ar+ sputtering and vacuum-annealing at 900 K.

These supported nanosheets were then calcined at 673 K under air at near atmospheric

pressure with a flow rate of 100 cm3/min for 6 h to remove the OSDA in the pores.

73

5.2.2 Bilayer Silicate

A Ru(0001) single crystal surface was cleaned by several cycles of Ar+

sputtering followed by annealing at 1250 K. The surface was then exposed to 4×10-6

mbar O2 at 1200 K to form a chemisorbed (2×2)-3O/Ru(0001) overlayer. The bilayer

silicate film was grown on the (2×2)-3O/Ru(0001) surface as described elsewhere.43,44

Briefly, Si was thermally evaporated onto the (2×2)-3O/Ru(0001) surface at room

temperature and 2.7×10-7 mbar O2. Next, the film underwent crystallization at 1200 K

in 4×10-6 mbar O2 for 10 min and subsequent cooling in O2 (at the same pressure). The

resulting film was a polymorphous silicate, containing both the bilayer structure and

amorphous silica particles on top of the bilayer.

5.2.3 Scanning Electron and Atomic Force Microscopies (SEM and AFM)

SEM and AFM provide information on surface topography and roughness,

respectively. SEM images were acquired as described by Kim et al.154 AFM images

were acquired in tapping-mode at room temperature in a 5 µm × 5 µm region.

5.2.4 Infrared Reflection-Absorption Spectroscopy (IRRAS)

IRRAS provides information on phonon vibrations for the framework of

porous materials and adsorbed molecules within the pores. The instrument uses a

Bruker Vertex 80V spectrometer from which infrared light is directed into a UHV

analysis chamber to interact with the sample surface at a grazing incidence (~8°). The

reflected light is then collected on a liquid-nitrogen cooled mercury-cadmium-telluride

(MCT) detector located in an external chamber. Spectra were collected for

wavelengths in the range of 700 – 5000 cm-1. The analysis chamber has KBr windows

to allow IR light to enter and exit the sample compartment. The beam path is in

vacuum both between the IR source and analysis chamber entry port and between the

74

analysis chamber exit port at the MCT detector. The system incorporates a polarizer,

which can easily tune the IR light interacting with the sample from p-polarized to s-

polarized. In this work, the s-polarized spectrum was collected and subtracted from the

p-polarized spectrum for all experiments to eliminate gas-phase contributions when

working at elevated pressures. The samples were mounted on a stainless-steel flag-

type holder with a K-type thermocouple allowing for in situ temperature readings. The

sample was heated with a halogen lamp located on the back of the sample holder.

Before performing IRRAS, the samples were cleaned by annealing in 1×10-5 mbar O2

at 750 K for 5 min to remove surface contaminates. During the IRRAS measurements,

the temperature of the sample was maintained at either 300, 473, or 573 K. Millipore

DI water, D2O (>99 atom %, Sigma Aldrich), or H218O (97 atom %, Sigma Aldrich)

was held in a glass cylinder and underwent freeze-pump-thaw cycles to remove the

dissolved gases prior to entering the IRRAS analysis chamber. Water vapor was dosed

up to 3 mbar with a leak valve. All of the spectra underwent baseline correction using

the Bruker OPUS Version 7.5 software. The SiOH peak areas were obtained using the

Peak Analyzer tool in Origin 2019.

5.2.5 X-ray Photoelectron Spectroscopy (XPS)

XPS provides information on the chemical composition and the oxidation

states of all the elements in a sample. XPS experiments were performed using a

system equipped with a SPECS PHOIBOS NAP 150 hemispherical analyzer and a

monochromatic Al Kα X-ray source. Spectra were acquired at room temperature and

under UHV conditions (2×10-9 mbar). The spectra were acquired on sample areas

smaller than 300 µm × 300 µm and were processed in the CasaXPS Version

2.3.19PR1.0 software.

75


5.3.1 IRRAS Characterization of 2-D Siliceous Nanomaterials

Prior to hydrothermal conditions, the 2-D siliceous nanomaterials were

characterized by IRRAS. Crystalline materials show phonon framework vibrations

within the mid-IR range, i.e., 4000 – 400 cm-1.155 In the MFI nanosheets supported on

Au(111), we observed a high frequency (HF) mode of 1247 cm-1 and a low frequency

(LF) mode of 1174 cm-1 (Figure 5-2a), which are consistent with measurements shown

in a previous report.156 Kestell et al. previously suggested that the HF mode likely

corresponds to Si-O-Si linkages between the pores, while the LF mode corresponds to

internal linkages within the pores.156 The peak position at 3724 cm-1 (Figure 5-2c)

corresponds to SiOH groups. Figure D-1 shows the SEM and AFM images of the MFI

nanosheets after deposition, where most of the surface is covered by individual

nanosheets, with some regions showing the overlap between adjacent nanosheets. The

AFM image reveals the average nanosheet thickness is 7 nm, consistent with Figure

5-1a.

The bilayer silicate supported on Ru(0001) revealed a sharp band at 1287 cm-1,

characteristic of a bilayer structure, and a broad region centered at 1182 cm-1,

corresponding to amorphous silica (Figure 5-2b). Amorphous silica can be produced

in 2-D materials by depositing Si in excess to that which is needed for the bilayer.44

We want to emphasize that by amorphous silica we refer here to 3-D disordered silica,

and not to the vitreous form bilayer silica, which is indistinguishable from the

crystalline one by IR. As shown in Figure 5-1b, a perfectly crystalline bilayer would

not contain amorphous silica or SiOH groups; however, our bilayer contains some

76

amorphous regions, which likely give rise to SiOH groups.46 We observed SiOH

groups in our bilayer silicate at 3736 cm-1 (Figure 5-2c).

The difference in SiOH peak positions for the MFI nanosheets (3724 cm-1) and

the polymorphous bilayer silicate (3736 cm-1) was likely because of SiOH’s

orientation within each material. Previous IR studies on silica assigned features from

3720 –3730 cm-1 to isolated, external SiOH groups,157–159 whereas isolated, internal

SiOH groups were assigned to 3735 – 3740 cm-1.160 As shown in Figure 5-1a, SiOH

groups are expected to terminate the truncated MFI framework. In the polymorphous

bilayer silicate, the SiOH groups could arise either from defects in the bilayer surface

or from external SiOH in the amorphous silica particles on top of the bilayer. Given

that the bilayer is a covalently saturated framework, it can be thought of as the wall of

an infinitely large pore. In this scenario, a SiOH group arising from a defect in the

structure would behave like an internal SiOH group, and this is consistent with the

peak we see at 3736 cm-1. In studies by Zhuravlev on amorphous silica, isolated SiOH

Figure 5-2: IRRAS spectra for the SiO2 phonons in (a) MFI nanosheets supported on

Au(111) and (b) bilayer silicate supported on Ru(0001). The SiOH regions for the

MFI nanosheets and bilayer silicate are provided in (c). Scale bars represent the scale

of the y-axis.

35003600370038003900

Ref

lect

an

ce

Wavenumber (cm-1)

10001100120013001400R

efle

ctan

ce

Wavenumber (cm-1)

10001100120013001400

Ref

lect

an

ce

Wavenumber (cm-1)

(b)(a) (c)

5% 0.5%1247 cm-1

1174 cm-1

1182 cm-11287 cm-1

Bilayer

MFI

3724 cm-1

3736 cm-10.05%

MFI Bilayer

77

groups also showed vibrational frequencies in the same range as internal SiOH groups

in crystalline porous silicates.161

In the following sections, we describe the results for applying hydrothermal

conditions to the MFI nanosheets and the bilayer silicate containing an amorphous

region. Throughout this report, we use the terms “bilayer” and “bilayer silicate”

interchangeably to reference the polymorphous bilayer silicate. Note the other regions

of the IRRAS for the MFI nanosheets and the bilayer silicate did not contain any

features and are therefore not shown. XPS spectra for the MFI nanosheets and the

bilayer silicate are provided in Figure D-2 and Figure D-3, respectively.

5.3.2 Progressive Hydrothermal Applications to 2-D Nanomaterials

Prior studies have shown when water interacts with a silica-containing

framework, Si-O-Si linkages undergo hydroxylation to form SiOH.53–55 An example of

the hydrolysis reaction to form two SiOH groups is shown in Eq. 5-1.54

Si-O-Si + H2O ↔ SiOH + SiOH Eq. 5-2

Given this reaction, we began by monitoring changes in SiOH formation in the

MFI nanosheets and the bilayer silicate in situ using IRRAS (Figure 5-3). We

calculated changes in SiOH formation by taking the ratio of the peak area obtained in

reflectance mode for SiOH during the hydrothermal condition relative to the peak area

obtained prior to treatment at 300 K and under UHV conditions. We denoted the ratio

as SiOH:SiOH300K, UHV throughout this work. The peak area was determined by taking

the area of the peak between the flat baselines on either side of the peak. In the silanol

region (3500 – 3900 cm-1), we found the peak areas obtained in the absorbance- and

reflectance-modes followed the same quantitative trends (Figure D-4). Therefore, for

the remainder of this manuscript, we have calculated the ratio of SiOH:SiOH300K, UHV

78

using the reflectance-mode spectra. Hydrothermal conditions were applied

progressively to the same sample (by increasing temperature and/or pressure). We

emphasize that the sample was not cooled down to 300 K and placed under UHV

conditions between each experiment.

Starting with the MFI nanosheets, the first spectrum shown in Figure 5-3a was

acquired prior to hydrothermal conditions. The first bar in the bar graph (Figure 5-3b)

corresponds to the reference spectrum, and thus it has a ratio of SiOH:SiOH300K,UHV

that is equal to 1. The error bars represent the noise associated with the measurement

and were obtained by taking the average peak area of the noise along the baseline.

When discussing the ratio of SiOH:SiOH300K,UHV throughout this manuscript, we

provide the average value. Proceeding on to progressive hydrothermal conditions, at

473 K and 1×10-3 mbar H2O for 60 s, the ratio of SiOH:SiOH300K, UHV remained

centered around 0.9, within error of the reference, meaning no change in SiOH

formation was observed (Figure 5-3a, Figure 5-3b). In the subsequent experiments, the

pressure was varied between 1×10-2, 1×10-1, and 1 mbar H2O each for 1 h at constant

temperature (473 K). The ratio of SiOH:SiOH300K, UHV remained comparable, implying

these hydrothermal conditions did not affect SiOH formation. Only when the pressure

was increased to 3 mbar H2O at 473 K, did the ratio of SiOH:SiOH300K, UHV begin to

change from 1 to 1.7. In the final hydrothermal condition applied at 573 K and 3 mbar

H2O, which was the highest temperature and H2O pressure that could be investigated

here, the ratio of SiOH:SiOH300K, UHV went up slightly more, from 1.7 to 1.8. After

H2O evacuation, the sample was cooled down to 300 K, and the area of the OH peak

from MFI nanosheets reverted to the initial state. In addition, the Si-O-Si linkages

from the extended framework remained intact, as indicated by the unchanged phonon

79

vibrations in IRRAS spectra shown in Figure D-5. Note there was a shift in the SiO2

phonons during hydrothermal conditions, but this was due to temperature change, no

H2O; we address the effects of temperature in more detail later in this Chapter (Figure

5-6.

Fully reversible hydrolysis and lack of framework degradation observed here

were also shown for a bulk zeolite in work by Heard et al.54 In that study, the zeolite

chabazite (CHA) was immersed in room temperature oxygen-17 labeled water

(H217O).54 The CHA framework underwent exchange from Si-16O-Si to Si-17O-Si.

Later in this Chapter, we will address isotopic labeling studies, which were performed

to gain insights about the location of the new SiOH groups formed in the MFI

framework. We want to emphasize that, even though our work and the work of Heard

et al. both show reversible hydrolysis without framework degradation, the materials

and the conditions are very different.54 In this work, we use thin (nm scale) films of

pure SiO2 while in Heard et al.’s work, the material studied is an aluminosilicate CHA

structure (a bulk material, which also contains Al in the framework). Additionally, in

our work, the experiments are carried out in an ultra-high vacuum chamber (UHV),

with controlled exposure to water vapor at low relative humidity, while in Heard et

al.’s work the experiments are carried out under liquid conditions.

For the bilayer silicate, hydrothermal conditions did not increase the number of

SiOH groups in the bilayer silicate. The ratio of SiOH:SiOH300K, UHV remained

centered at ~1 for all treatments (Figure 5-3c, Figure 5-3d). These results were not

surprising, as prior works suggest the bilayer silicate is a durable surface that requires

extreme treatment to undergo any form of degradation. For example, in a previous

study by Yang et al., only small amounts of SiOH groups could be formed in the

80

bilayer silicate upon deposition of an ice-like film of water at 100 K and subsequent

heating to room temperature.55 In a later study, Yu et al. found that the number of

SiOH groups increased only when an ice-like water film on the bilayer silicate was

exposed to an electron beam at cryogenic conditions and then heated to room

temperature.56 The formation of more SiOH groups by electron stimulation came at

the expense of partial, irreversible destruction to the bilayer’s crystalline framework.56

Dissolution of the bilayer silicate was also observed by Kaden et al. under basic

(pH>10) and highly acidic conditions (pH<1), both at room temperature and 363 K.53

Consistently, in our work, we did not observe any degradation in the crystalline

framework of the bilayer silicate when applying these mild hydrothermal conditions

(Figure D-5). We note that the bilayer silicates are comprised of an extended

framework of hexagonal prisms; however, these silicates do not have pores like

zeolites do, so the possibility of “internal” silanol groups can be discarded.

The above conclusions remain the same in experiments using D2O (Figure D-

6). Initially, the MFI nanosheets and bilayer silicate contained SiOH. Upon applying

hydrothermal conditions at 473 K and low pressures of D2O, we observed an exchange

from SiOH to SiOD. At 573 K and 3 mbar D2O, the peak area for SiOD (2740 cm-1) in

the MFI nanosheets grew substantially larger compared to all other treatments and

decreased in size once all D2O had been evacuated. The bilayer silicate maintained the

same peak area for SiOD (2760 cm-1) at the most extreme condition and upon removal

of D2O. In summary, the 2-D nanomaterials behaved similarly when exposed to H2O

and D2O. Namely, the formation of SiOH/SiOD increased at 573 K and 3 mbar

H2O/D2O in the MFI nanosheets, and the effects were fully reversible. In contrast, the

formation of new SiOH/SiOD groups did not occur in the bilayer silicate.

81

In summary, new SiOH groups form at mild temperatures and pressures on the

MFI nanosheets, but not the bilayer silicate. The reason behind the differences

observed between both materials could be related to higher thermodynamic stability of

the bilayer silicate compared to the MFI nanosheets. Additionally, the bilayer silicate

is a denser, self-containing framework, that has no dangling bonds, i.e., if there are no

defects in the structure, all bonds are covalently saturated. In contrast, MFI nanosheets

are less dense, truncated versions of a 3-D framework, whose exposed surface needs to

reconstruct or hydroxylate in the truncation plane. A way to illustrate this further is by

considering the higher thermal stability of bilayer silicates (synthesized at 1200 K)

compared to MFI nanosheets, which exhibit changes in morphology at temperatures

above 1000 K, or lower in the presence of steam.37

82

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

Ra

tio

of

SiO

H:S

iOH

30

0K

, U

HV p

eak

are

as

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

Ra

tio

of

SiO

H:S

iOH

30

0K

, U

HV p

eak

are

as

35003600370038003900

Ref

lect

an

ce

Wavenumber (cm-1)

3736 cm-1

35003600370038003900

Ref

lect

an

ce

Wavenumber (cm-1)

3724 cm-1

(b)

0.10%

300 K

473 K

573 K

300 K

300 K

473 K

573 K

300 K

0.20%

(a)

(c) (d)

473 K

473 K

473 K

473 K

473 K

473 K

473 K

473 K

Figure 5-3: (a) IRRAS of MFI nanosheets supported on Au(111) and (b) quantification

of the IRRAS peak areas for SiOH, taken relative to the initial amount of SiOH

obtained prior to hydrothermal conditions at 300 K and UHV conditions. (c) IRRAS

of bilayer silicate supported on Ru(0001) and (d) quantification of the IRRAS peak

areas for SiOH. All spectra were taken in the presence of H2O and after 1 h, with the

exception of the pre-/post-H2O and 473 K, 1x10-3 mbar H2O treatments, which had

been exposed to the condition for 60 s and were not in the presence of H2O. The

conditions of each acquisition in (a) and (c) are provided in the x-axes of (b) and (d),

respectively.

83

5.3.3 Single-Step Hydrothermal Application (573 K, 3 mbar H2O, 1 h)

As shown in the previous section, an increase in SiOH formation in the MFI

nanosheets was observed at 473 and 573 K and 3 mbar H2O; however, multiple

attempts of milder hydrothermal treatments had been applied to the sample prior to

these treatments. This section addresses the effect of hydrothermal conditions at only

the most severe hydrothermal conditions ( i.e., at 573 K and 3 mbar H2O for 1 h),

reached in a single step. The data presented in Figure 5-4 is on the same sample of

MFI nanosheets mentioned previously, but the sample has been regenerated as

described in the Methods section. Quantification of SiOH was performed in the same

way as described in the Methods section; the SiOH peak areas were taken relative to

the peak area for SiOH obtained prior to hydrothermal conditions at 300 K and UHV

in the top-most spectrum of Figure 5-4a. The first bar in Figure 5-4b is the reference

point (ratio of SiOH:SiOH300K, UHV equal to 1). Next, a spectrum was acquired at 573

K and under UHV. The ratio of SiOH:SiOH300K, UHV remained at ~ 1, implying no

SiOH changes result from increasing the temperature. After increasing the pressure of

H2O to 3 mbar (still at 573 K), the ratio of SiOH:SiOH300K, UHV increased from 1 to

1.7. Upon cooling the sample and evacuating water, the ratio dropped to 1.2. Overall,

the results were consistent with those obtained during progressive hydrothermal

treatments. The data suggests that the increase in the number of SiOH groups is the

same when the most severe hydrothermal condition is applied in a single step, or

progressively to reach the same condition. The bilayer silicate did not experience any

new SiOH formation, consistent with the results shown after progressive hydrothermal

treatment, as presented in Figure D-7.

84

5.3.4 Mechanistic Insights from H218O

We reported an increase in SiOH groups under 3 mbar H2O at 473 and 573 K,

whether the MFI nanosheets had undergone progressive hydrothermal treatments

(Figure 5-3), or the same conditions were reached in a single step (Figure 5-4a). An

interesting observation, when comparing the spectra at 300 K in UHV before and after

hydrothermal treatments, is that the peaks for the SiOH groups are sharper after the

treatment. While we are not certain of the origin of this, a possible explanation could

be related to heterogeneities in the distribution and proximity of the hydroxyl groups

on the surface in the initial structures. The mild hydrothermal treatment may lead to

reorganization of the original SiOH groups, leading to a more homogeneous surface

distribution. Furthermore, as shown in Figure 5-3a, during extreme hydrothermal

conditions, we observed a peak for isolated, external SiOH groups (3724 cm-1) as well

Figure 5-4: (a) IRRAS of MFI nanosheets supported on Au(111). The effect of H2O on

SiOH formation in MFI nanosheets supported on Au(111). Pre-, post-, 573 K under

UHV, and 573 K under 3 mbar H2O for 1 h are specified. (b) Quantification of the

IRRAS peak areas for SiOH, taken relative to the initial amount of SiOH obtained

prior to hydrothermal conditions at 300 K and UHV conditions.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

Ra

tio

of

SiO

H:S

iOH

30

0K

, U

HV p

eak

are

as

35003600370038003900

Ref

lect

an

ce

Wavenumber (cm-1)

3724 cm-1

(b)(a)

300 K, UHV

573 K, UHV

573 K, 3 mbar

300 K, UHV

0.20%

Pre-H2O

Post-H2O

85

as a broad region extending to ~3620 cm-1, which likely corresponds to hydrogen-

bonded SiOH groups.158,159 In a previous work, Karbowiak et al. found that hydrogen-

bonded SiOH groups formed in silicalite-1, a pure silica MFI-type zeolite, when

exposed to pressures between 80 and 100 MPa H2O followed by evacuation of H2O at

room temperature.159 Hydrogen-bonded SiOH groups were also observed post-H2O in

Figure 5-4a, though not in Figure 5-3a. The presence of these hydrogen-bonded SiOH

groups in one case but not the other was most likely because H2O had been evacuated

upon cooling down to 300 K and trace amount of water remained (Figure 5-4a). In

contrast, in Figure 5-3a, H2O was evacuated at elevated temperature and then cooled.

Note we included both isolated and hydrogen-bonded SiOH groups in the peak area

for SiOH (Figure 5-3b and Figure 5-4b); however, these SiOH groups have different

extinction coefficients, making it difficult to deduce quantitative comparisons between

the two.162 Qualitatively, however, the number of all SiOH groups increases upon

exposure to H2O at these mild hydrothermal conditions

To acquire insights about the location of new SiOH groups, we exposed the

MFI nanosheets to oxygen-18 labeled water (H218O). H2

18O can be used to determine

if an isotopic exchange occurs between H218O and the internal Si-16O-Si framework in

IRRAS. In Figure 5-5, we have proposed a mechanism in which H218O cleaves the Si-

16O-Si linkage to form two hydrogen-bonded SiOH groups.

Figure 5-5: A possible reaction pathway showing the formation of SiOH in the MFI

nanosheets with H218O.

86

Upon condensation of SiOH groups to reform the Si-O-Si siloxane linkage, there is a

50% chance that oxygen-18 replaces the original oxygen-16. If this exchange occurs in

the MFI framework rather than on the defective sites, the framework phonon

frequencies are expected to red-shit as a result of the heavier 18O isotopes in the Si-

18O-Si linkages (if sufficient 16O/18O exchange events take place).53,55,56 Defect sites

encompass regions in the structure that deviate from the perfect 3-D MFI framework.

These sites include the SiOH groups (evidenced by the 3724 cm-1 feature in the IR

spectra) at the truncation plane of the of the framework (i.e., the surface of the

nanosheet). Note that the SiOH groups on the external surface of a nanosheet are

equivalent to the external SiOH groups on the surface of a bulk zeolite crystallite, and

the observed frequency at 3724 cm-1 is in agreement with this.

As we showed previously, applying the extreme hydrothermal condition (573

K and 3 mbar H2O) for 1 h was sufficient time to detect new SiOH groups in MFI

nanosheets. Therefore, we performed the hydrothermal condition for 1 h with H218O.

Although no changes were expected to occur in the bilayer silicate, experiments with

H218O were performed for comparison to the MFI nanosheets. Figure 5-6 shows

consistent findings for both 2-D materials. The phonon frequencies corresponding to

the MFI framework were identical pre- and post-H218O. At the elevated temperature

(573 K, UHV), the phonons shifted to lower frequencies (due to thermal effects)163 but

remained unchanged upon addition of water (573 K, 3 mbar H218O). The fact that

there are no changes in the frequencies of the framework phonon vibrations after

applying the hydrothermal condition and evacuation indicates that the formation of Si-

18O-Si in the MFI framework does not occur to a detectable degree. Then, the

formation of additional SiOH groups under hydrothermal conditions must take place

87

on defect sites that are not part of the extended framework. We want to emphasize that

the fact that we do see additional SiOH groups by IR, but that no phonon vibration

shifts are observed upon using isotopically labeled water (with 18O), suggests that

these additional OH groups are not part of the extended framework that gives rise to

the phonon vibrations, or that the reversible hydroxylation does not take place in this

part of the structure to a detectable level.

We also investigated the effect of H218O on the SiOH groups. The exchange

from Si16OH to Si18OH only produced a difference of 10 cm-1 in the vibrational

frequency of OH groups, consistent with the literature,164–166 making it difficult to

differentiate the two peaks (with and without 18O). The Si-O-Si linkages in the bilayer

silicate remained intact and did not interact with water to form SiOH groups.

Figure 5-6: The effect of H218O on the crystalline structure of (a) MFI nanosheets

supported on Au(111) and (b) bilayer silicate on Ru(0001). Pre-, post-, 573 K under

UHV, and 573 K under 3 mbar H218O are specified.

10001100120013001400

Ref

lect

an

ce

Wavenumber (cm-1)

10001100120013001400

Ref

lect

an

ce

Wavenumber (cm-1)

573 K, UHV

Pre-H218O

573 K, 3 mbar H218O

Post-H218O

0.5%

573 K, UHV

(b)(a)

573 K, 3 mbar H218O

Pre-H218O Post-H2

18O

5%

88

5.4 Conclusions

In this work, in situ IRRAS studies revealed the effects of mild hydrothermal

conditions on 2-D model systems, namely all-Si MFI nanosheets supported on

Au(111) and a polymorphous bilayer silicate supported on Ru(0001). We were able to

detect changes in SiOH formation, which were previously difficult to study using bulk

techniques, such NMR and XRD. In particular, we found that SiOH groups increased

in the MFI nanosheets at 473 and 573 K under a H2O pressure of 3 mbar, but were

unaffected in the bilayer silicate at the same conditions. Furthermore, the effects in the

MFI nanosheets were the same if these mild hydrothermal conditions were reached

progressively or in one step. A fraction of the SiOH groups formed in the MFI

nanosheets were hydrogen-bonded and were deduced to form in defect sites, not in the

extended framework. In the MFI nanosheets, the effects of mild hydrothermal

conditions were fully reversible and did not result in framework degradation. The

results shown here suggest that well-ordered, porous materials, inherently terminated

in SiOH groups, such as the MFI nanosheets, are susceptible to an increase in SiOH

groups under mild hydrothermal conditions. In contrast, covalently saturated ones,

such as bilayer silicates, which lack SiOH groups in their ideal structure, are not

affected. These conclusions improve our fundamental understanding of the effects of

mild hydrothermal conditions on 2-D siliceous nanomaterials and may aid in

understanding the impact of these conditions on 3-D ones.

5.5 Acknowledgements

This work was financially supported by the Catalysis Center for Energy

Innovation, an Energy Frontier Research Center funded by the U.S. Department of

Energy, Office of Science, Office of Basic Energy Sciences under Award Number

89

DE-SC0001004. Research is carried out in part at the Center for Functional

Nanomaterials, which is a U.S. DOE Office of Science Facility at Brookhaven

National Laboratory under Contract No. DE-SC0012704. The authors thank John S.

Ruano-Salguero for his helpful discussions.

90

CONCLUSIONS AND FUTURE WORK

Lignocellulosic biomass offers a promising alternative feedstock to petroleum,

but its high oxygen content is one of the most significant barriers for its conversion

into low-oxygen containing fuels and chemicals traditionally derived from petroleum.

This thesis focuses on the catalytic systems and processes involved in biomass

upgrading, including: (1) an investigation into the active species of the metal salt,

AlCl3, in glucose to fructose isomerization, (2) the development of a catalytic process

for the production of a bio-lubricant base oil from biomass-derived platform

chemicals, and (3) a fundamental understanding on the effects of mild hydrothermal

conditions experienced during catalyst regeneration on siliceous nanomaterials. Such

knowledge will aid in the development of improved catalysts and processes to obtain

products from biomass. This section provides a holistic set of conclusions and their

impact on future research directions.

6.1 Thesis Summary

In Chapter 2, a review of the literature on the critical role metal salts play in

glucose conversion was provided. It was shown that metal salts affect glucose ring

opening, and their speciation facilitates a C2 → C1 hydride transfer to produce

fructose, but detecting the active metal species responsible for glucose conversion is

challenging.

Chapter 6

91

Chapter 3 showed that the Al species’ concentrations can be deduced by

combining experiments, modeling, and glucose isomerization kinetics. It was found

that 27Al NMR detected the dominant Al(H2O)63+ complex. These measurements

combined with pH and equilibrium expressions for the aluminum hydrolysis reactions

were used to estimate the concentrations of the [Al(H2O)4(OH)2]1+ and

[Al(H2O)5(OH)]2+ species. Aluminum nanoparticles were quantified through a

combination of ICP-MS, DLS, and ultrafiltration. The concentrations of the aqueous

and solid aluminum species were correlated to the glucose isomerization rate. At

sufficiently high temperatures, the hydrolyzed [Al(H2O)4(OH)2]1+ species scales

linearly with the glucose isomerization rate, indicating this complex is the active one.

In Chapter 4, a strategy was developed to synthesize branched alkanes for

lubricant base oil in two steps from 12-tricosanone, which can be obtained from

bioderived fatty acids, and furfural, which can be obtained from lignocellulosic

biomass. The reaction pathway involved carbon-carbon coupling through aldol

condensation followed by HDO. Various solvents (non-polar, aprotic and polar,

protic) and reaction conditions were screened to achieve a maximum yield of 94.3% of

aldol condensation products, containing the majority of a C33 furan (79.5%) followed

by a C28 furan (14.8%). Subsequent HDO of aldol condensation products over an Ir-

ReOx/SiO2 catalyst produced lubricant-ranged branched alkanes (C28 and C33) with

61.4 % yield and small fractions (< 11%) of alkanes with carbon numbers between C15

and C10. The viscous properties of the produced bio-lubricant base oil were

comparable to commercial petroleum-derived base oils. This approach serves as a

potential stepping-stone to replace petroleum-derived base oils and, in turn, reduce

greenhouse gas emissions associated with current lubricant production.

92

In Chapter 5, the effects of mild hydrothermal conditions, in situ, on siliceous

nanomaterials, namely all-Si MFI nanosheets supported on Au(111) and a

polymorphous bilayer silicate supported on Ru(0001), were observed by IRRAS. It

was found that mild hydrothermal conditions, i.e., 473 and 573 K and 3 mbar H2O,

increase the formation of SiOH groups in the MFI nanosheets, but do not change the

polymorphous bilayer silicate. The effects of mild hydrothermal conditions were

reversible and did not result in framework degradation. Implications provided in this

work provided a fundamental understanding on the effects of mild hydrothermal

conditions on 2-D siliceous nanomaterials and serve as a starting point when

considering these effects on 3-D ones.

6.2 Future Work

6.2.1 Metal Salt Catalyzed Glucose Isomerization

Chapter 3 of this thesis shows that the partially hydrolyzed aluminum ion

[Al(H2O)4(OH)2]1+ is the most active species for glucose to fructose isomerization in

water. It is shown that sufficient time is necessary to prevent changes in metal salt

speciation during glucose conversion. The time for equilibrating the salt species

depends on the metal salt and temperature, and a simple way to estimate it is by

measuring signatures, such as pH vs. time, prior to adding reagents. Future

experiments require understanding how much preheating time is necessary for other

highly effective metal salts in glucose isomerization, such as CrCl3. In addition, future

investigation is necessary to understand the effects of preheating on the glucose-

fructose reaction rate.

93

Anchoring the active species on a support to prevent agglomeration may be

worth pursuing in the future. In addition, more work is needed to delineate the

similarities and differences between homogeneous and heterogeneous catalysts studied

thus far in glucose isomerization. For example, prior work found that SnCl4 is not as

active as Sn-BEA in the isomerization reaction, even though both contained Sn.

Therefore, there is a need to understand the influence of the framework on Lewis

acidity of the metal center, potentially by ab initio calculations. In addition, the effects

of the size and structure of the reactant compared to the pore size and shape of the

heterogeneous catalyst require better understanding.

6.2.2 Bio-Lubricant Base Oil Production Through Aldol Condensation

Chapter 4 of this thesis develops a strategy to synthesize branched alkanes for

lubricant base oil in two steps from 12-tricosanone, obtained from bioderived fatty

acids, and furfural, obtained from lignocellulosic biomass. The reaction pathway

involves carbon-carbon coupling through aldol condensation followed by

hydrodeoxygenation (HDO). While reaction parameters, such as time, solvent, and

ratio of reagents in aldol condensation, are studied, additional parameters, including

feedstocks, catalysts, and separation procedures should be considered.

Other fatty acids to derive long-chain ketones for aldol condensation are

recommended. In the current strategy, the ketone, 12-tricosanone, is derived from

lauric acid, a fatty acid naturally found in coconut and palm kernel oils. Although

these oils are renewable, growing concerns indicate that the increase in the demand for

products derived from coconut and palm kernel oils could lead to the mass

deforestation of rainforests. If possible, other sources of fatty acids, especially those

that may still be present in waste cooking oils, should be considered.

94

There is also potential to explore different catalysts for the aldol condensation

and HDO steps. Solid organic bases to consider for aldol condensation include Al2O3,

TiO2, MgO, CaO, and KCN. In the HDO reaction, additional solid acid supported

metal-based catalysts should be explored as alternatives to Ir-ReOx/SiO2, since Ir and

Re are rare elements. These catalysts are industrially available and include Pd/C,

Pd/SiO2, and Pt/C, among others. In addition, other promising HDO catalysts of the

form 1M2MO/SiO2 will likely be effective. Combinations to consider include 1M = Ru,

Ni, Co, Pd, Pt, or Rh and 2M = Mo, W, Nb, Mn, V, Ce, Cr, Zn, Co, Y, or Al.

Currently, the products obtained from aldol condensation are separated from

the catalyst through extraction into dichloromethane (CH2Cl2). CH2Cl2 is then

removed by rotary evaporation. As the extraction solvent, CH2Cl2, possesses health

risks, other extraction solvents should be considered, or potentially eliminated,

depending upon catalyst selection.

6.2.3 Hydrothermal Conditions Applied to 2-D Siliceous Nanomaterials

Chapter 5 of this thesis studies the effect of mild hydrothermal conditions on 2-

D siliceous nanomaterials, namely MFI nanosheets supported on Au(111) and a

polymorphous bilayer silicate supported on Ru(0001). It is shown that SiOH groups

form at elevated temperatures and pressure of H2O, i.e., 473 and 573 K and 3 mbar

H2O, in the MFI nanosheets, but not in the polymorphous bilayer silicate. The effects

are reversible and do not result in framework degradation.

SiOH groups serve as anchoring points for various chemical species. In the

future, it is recommended that experiments target understanding how metals anchor to

siliceous supports, as this could aid in catalysts design. Appendix E highlights initial

attempts to anchor phosphorus species onto a crystalline bilayer silicate. It is shown

95

that phosphorus species desorb from the surface, likely because there are not any

SiOH anchoring points in a crystalline bilayer. In the future, it is recommended to

explore how phosphorus species interact with surfaces containing SiOH groups. Mild

hydrothermal treatment is one way to incorporate more SiOH groups onto MFI

nanosheets, but the effects are reversible, only being observed during the hydrothermal

treatment. Therefore, future studies should attempt dosing phosphorus species while

applying the hydrothermal treatment, or developing permanent SiOH groups through

more intense hydrothermal conditions than those tested previously. Currently, the

maximum condition that can be implemented in the IRRAS chamber is 573 K and 3

mbar H2O. A new experimental set-up, such as an autoclave, is worth attempting to

test hydrothermal conditions beyond the current maximum. It is suggested that the

material be examined by IRRAS before and after the hydrothermal treatment to

determine if irreversible SiOH formation has occurred. It may also be worthwhile to

consider dosing the phosphorus species in an external chamber to avoid contaminating

the IRRAS chamber.

96

(1) Langholtz, B. M. H. J.; Stokes, L. M.; Eaton. 2016 BILLION-TON REPORT

Advancing Domestic Resources for a Thriving Bioeconomy. Oak Ridge Natl.

Lab. 2016.

(2) Mamman, A. S.; Lee, J.-M.; Kim, Y.-C.; Hwang, I. T.; Park, N.-J.; Hwang, Y.

K.; Chang, J.-S.; Hwang, J.-S. Furfural: Hemicellulose/Xylosederived

Biochemical. Biofuels, Bioprod. Biorefining 2008, 2, 438–454.

(3) Alonso, D. M.; Bond, J. Q.; Dumesic, J. A. Catalytic Conversion of Biomass to

Biofuels. Green Chem. 2010, 12, 1493–1513.

(4) Matson, T. D.; Barta, K.; Iretskii, A. V.; Ford, P. C. One-Pot Catalytic

Conversion of Cellulose and of Woody Biomass Solids to Liquid Fuels. J. Am.

Chem. Soc. 2011.

(5) Dutta, S.; De, S.; Alam, M. I.; Abu-Omar, M. M.; Saha, B. Direct Conversion

of Cellulose and Lignocellulosic Biomass into Chemicals and Biofuel with

Metal Chloride Catalysts. J. Catal. 2012.

(6) Zhou, S.; You, K.; Yi, Z.; Liu, P.; Luo, H. Metal Salts with Highly

Electronegative Cations as Efficient Catalysts for the Liquid-Phase Nitration of

Benzene by NO 2 to Nitrobenzene.

(7) Choudhary, V.; Mushrif, S. H.; Ho, C.; Anderko, A.; Nikolakis, V.;

Marinkovic, N. S.; Frenkel, A. I.; Sandler, S. I.; Vlachos, D. G. Insights into the

Interplay of Lewis and Brønsted Acid Catalysts in Glucose and Fructose

Conversion to 5-(Hydroxymethyl)Furfural and Levulinic Acid in Aqueous

Media. J. Am. Chem. Soc. 2013, 135, 3997–4006.

(8) Cheng, Z.; Everhart, J. L.; Tsilomelekis, G.; Nikolakis, V.; Saha, B.; Vlachos,

D. G. Structural Analysis of Humins Formed in the Brønsted Acid Catalyzed

Dehydration of Fructose. Green Chem. 2018, 20, 997–1006.

(9) Teong, S. P.; Yi, G.; Zhang, Y. Hydroxymethylfurfural Production from

Bioresources: Past, Present and Future. Green Chem. 2014, 16, 2015–2026.

(10) Brownlee, H. J.; Miner, C. S. Industrial Development of Furfural. Ind. Eng.

Chem. 1948, 40, 201–204.

(11) Moliner, M.; Roman-Leshkov, Y.; Davis, M. E. Tin-Containing Zeolites Are

Highly Active Catalysts for the Isomerization of Glucose in Water. Proc. Natl.

Acad. Sci. 2010, 107, 6164–6168.

(12) van Putten, R.-J.; van der Waal, J. C.; de Jong, E.; Rasrendra, C. B.; Heeres, H.

J.; de Vries, J. G. Hydroxymethylfurfural, A Versatile Platform Chemical Made

from Renewable Resources. Chem. Rev. 2013, 113, 1499–1597.

(13) Zhao, H.; Holladay, J. E.; Brown, H.; Zhang, Z. C. Metal Chlorides in Ionic

REFERENCES

97

Liquid Solvents Convert Sugars to 5-Hydroxymethylfurfural. Science (80-. ).

2007, 316, 1597–1600.

(14) Choudhary, V.; Pinar, A. B.; Lobo, R. F.; Vlachos, D. G.; Sandler, S. I.

Comparison of Homogeneous and Heterogeneous Catalysts for Glucose-to-

Fructose Isomerization in Aqueous Media. ChemSusChem 2013, 6, 2369–2376.

(15) Tang, J.; Guo, X.; Zhu, L.; Hu, C. Mechanistic Study of Glucose-to-Fructose

Isomerization in Water Catalyzed by [Al(OH)<inf>2</Inf>(Aq)]+. ACS Catal.

2015, 5, 5097–5103.

(16) Stewart, I. I. Electrospray Mass Spectrometry: A Tool for Elemental Speciation.

Spectrochim. acta, Part B At. Spectrosc. 1999, 54, 1649–1695.

(17) American Petroleum Institute. Annex E- API Base Oil Interchangeaility

Guidelines for Passenger Car Motor Oils and Diesel Engine Oils; 2015.

(18) Brown, S. Base Oil Groups: Manufacture, Properties, and Performance.

Tribology & Lubrication Technology. April 2015, pp 32–35.

(19) Ray, S.; Rao, P. V. C.; Choudary, N. V. Poly-α-Olefin-Based Synthetic

Lubricants: A Short Review on Various Synthetic Routes. Lubr. Sci. 2012, 24,

23–44.

(20) ExxonMobil. ExxonMobil Basestocks 2018 Pulse Report

http://www.exxonmobil.com/en/basestocks/news-insights-and-

resources/industry-insights/2018-industry-pulse-report.

(21) Rudnick, L. R. Synthetics, Mineral Oils, and Bio-Based Lubricants: Chemistry

and Technology; 2005.

(22) Liu, S.; Josephson, T. R.; Athaley, A.; Chen, Q. P.; Norton, A.; Ierapetritou,

M.; Siepmann, J. I.; Saha, B.; Vlachos, D. G. Renewable Lubricants with

Tailored Molecular Architecture. Sci. Adv. 2019, 5, 1–8.

(23) Ji, H.; Zhang, X.; Tan, T. Preparation of a Water-Based Lubricant from

Lignocellulosic Biomass and Its Tribological Properties. Ind. Eng. Chem. Res.

2017, 56, 7858–7864.

(24) Ji, H.; Wang, B.; Zhang, X.; Tan, T. Synthesis of Levulinic Acid-Based Polyol

Ester and Its Influence on Tribological Behavior as a Potential Lubricant. RSC

Adv. 2015, 5, 100443–100451.

(25) Gu, M.; Xia, Q.; Liu, X.; Guo, Y.; Wang, Y. Synthesis of Renewable Lubricant

Alkanes from Biomass-Derived Platform Chemicals. ChemSusChem 2017, 10,

4102–4108.

(26) Corma, A.; Renz, M.; Schaverien, C. Coupling Fatty Acids by Ketonic

Decarboxylation Using Solid Catalysts for the Direct Production of Diesel,

Lubricants, and Chemicals. ChemSusChem 2008, 1, 739–741.

(27) Shylesh, S.; Gokhale, A. A.; Sun, K.; Grippo, A. M.; Jadhav, D.; Yeh, A.; Ho,

C. R.; Bell, A. T. Integrated Catalytic Sequences for Catalytic Upgrading of

Bio-Derived Carboxylic Acids to Fuels, Lubricants and Chemical Feedstocks.

Sustain. Energy Fuels 2017, 1, 1805–1809.

(28) Balakrishnan, M.; Sacia, E. R.; Sreekumar, S.; Gunbas, G.; Gokhale, A. A.;

Scown, C. D.; Toste, F. D.; Bell, A. T. Novel Pathways for Fuels and

98

Lubricants from Biomass Optimized Using Life-Cycle Greenhouse Gas

Assessment. Proc. Natl. Acad. Sci. 2015, 112, 7645–7649.

(29) Sacia, E. R.; Balakrishnan, M.; Deaner, M. H.; Goulas, K. A.; Toste, F. D.;

Bell, A. T. Highly Selective Condensation of Biomass-Derived Methyl Ketones

as a Source of Aviation Fuel. ChemSusChem 2015, 8, 1726–1736.

(30) U.S. Energy Information Administration. Petroleum & Other Liquids

https://www.eia.gov/dnav/pet/PET_PRI_GND_A_EPD2DXL0_PTE_DPGAL_

W.htm (accessed Jun 6, 2019).

(31) Wheeler, G. U.S. Base Oil Price Report. Lubes “N” Greases 2019, 2.

(32) Serrano-Ruiz, J. C.; Dumesic, J. A. Catalytic Routes for the Conversion of

Biomass into Liquid Hydrocarbon Transportation Fuels. Energy Environ. Sci.

2011, 4, 83–99.

(33) Corma, A.; Oliver-Tomas, B.; Renz, M.; Simakova, I. L. Conversion of

Levulinic Acid Derived Valeric Acid into a Liquid Transportation Fuel of the

Kerosene Type. J. Mol. Catal. A Chem. 2014, 388–389, 116–122.

(34) Corma, A.; De La Torre, O.; Renz, M. Production of High Quality Diesel from

Cellulose and Hemicellulose by the Sylvan Process: Catalysts and Process

Variables. Energy Environ. Sci. 2012, 5, 6328–6344.

(35) Anbarasan, P.; Baer, Z. C.; Sreekumar, S.; Gross, E.; Binder, J. B.; Blanch, H.

W.; Clark, D. S.; Dean Toste, F. Integration of Chemical Catalysis with

Extractive Fermentation to Produce Fuels. Nature 2012, 491, 235–239.

(36) Elyassi, B.; Zhang, X.; Tsapatsis, M. Long-Term Steam Stability of MWW

Structure Zeolites (MCM-22 and ITQ-1). Microporous Mesoporous Mater.

2014, 193, 134–144.

(37) Guefrachi, Y.; Sharma, G.; Xu, D.; Kumar, G.; Vinter, K. P.; Abdelrahman, O.

A.; Li, X.; Alhassan, S.; Dauenhauer, P. J.; Navrotsky, A.; et al. Steam-Induced

Coarsening of Single-Unit-Cell MFI Zeolite Nanosheets and Its Effect on

External Surface Brønsted Acid Catalysis. Angew. Chemie - Int. Ed. 2020, 59,

2–9.

(38) Paul, G.; Bisio, C.; Braschi, I.; Cossi, M.; Gatti, G.; Gianotti, E.; Marchese, L.

Combined Solid-State NMR, FT-IR and Computational Studies on Layered and

Porous Materials. Chem. Soc. Rev. 2018, 47, 5684–5739.

(39) Holder, C. F.; Schaak, R. E. Tutorial on Powder X-Ray Diffraction for

Characterizing Nanoscale Materials. ACS Nano 2019, 13, 7359–7365.

(40) Sulpizi, M.; Gaigeot, M. P.; Sprik, M. The Silica-Water Interface: How the

Silanols Determine the Surface Acidity and Modulate the Water Properties. J.

Chem. Theory Comput. 2012, 8, 1037–1047.

(41) Boscoboinik, J. A.; Shaikhutdinov, S. Exploring Zeolite Chemistry with the

Tools of Surface Science: Challenges, Opportunities, and Limitations. Catal.

Letters 2014, 144, 1987–1995.

(42) Löffler, D.; Uhlrich, J. J.; Baron, M.; Yang, B.; Yu, X.; Lichtenstein, L.;

Heinke, L.; Büchner, C.; Heyde, M.; Shaikhutdinov, S.; et al. Growth and

Structure of Crystalline Silica Sheet on Ru(0001). Phys. Rev. Lett. 2010, 105,

99

146104(1)-146104(4).

(43) Boscoboinik, J. A.; Yu, X.; Yang, B.; Fischer, F. D.; Wodarczyk, R.; Sierka,

M.; Shaikhutdinov, S.; Sauer, J.; Freund, H. J. Modeling Zeolites with Metal-

Supported Two-Dimensional Aluminosilicate Films. Angew. Chemie - Int. Ed.

2012, 51, 6005–6008.

(44) Yang, B.; Kaden, W. E.; Yu, X.; Boscoboinik, J. A.; Martynova, Y.;

Lichtenstein, L.; Heyde, M.; Sterrer, M.; Włodarczyk, R.; Sierka, M.; et al. Thin

Silica Films on Ru(0001): Monolayer, Bilayer and Three-Dimensional

Networks of [SiO 4] Tetrahedra. Phys. Chem. Chem. Phys. 2012, 14, 11344–

11351.

(45) Yu, X.; Yang, B.; Anibal Boscoboinik, J.; Shaikhutdinov, S.; Freund, H. J.

Support Effects on the Atomic Structure of Ultrathin Silica Films on Metals.

Appl. Phys. Lett. 2012, 100, 151608-1-151608–4.

(46) Goldsmith, B. R.; Peters, B.; Johnson, J. K.; Gates, B. C.; Scott, S. L. Beyond

Ordered Materials: Understanding Catalytic Sites on Amorphous Solids. ACS

Catal. 2017, 7, 7543–7557.

(47) Agrawal, K. V.; Topuz, B.; Jiang, Z.; Nguenkam, K.; Elyassi, B.; Francis, L. F.;

Tsapatsis, M.; Navarro, M. Solution-Processable Exfoliated Zeolite Nanosheets

Purified by Density Gradient Centrifugation. AIChE J. 2013, 59, 3458–3467.

(48) Varoon, K.; Zhang, X.; Elyassi, B.; Brewer, D. D.; Gettel, M.; Kumar, S.; Lee,

J. A.; Maheshwari, S.; Mittal, A.; Sung, C. Y.; et al. Dispersible Exfoliated

Zeolite Nanosheets and Their Application as a Selective Membrane. Science

(80-. ). 2011, 334, 72–75.

(49) Jeon, M. Y.; Kim, D.; Kumar, P.; Lee, P. S.; Rangnekar, N.; Bai, P.; Shete, M.;

Elyassi, B.; Lee, H. S.; Narasimharao, K.; et al. Ultra-Selective High-Flux

Membranes from Directly Synthesized Zeolite Nanosheets. Nature 2017, 543,

690–694.

(50) Cho, H. J.; Ren, L.; Vattipalli, V.; Yeh, Y. H.; Gould, N.; Xu, B.; Gorte, R. J.;

Lobo, R.; Dauenhauer, P. J.; Tsapatsis, M.; et al. Renewable P-Xylene from

2,5-Dimethylfuran and Ethylene Using Phosphorus-Containing Zeolite

Catalysts. ChemCatChem 2017, 9, 398–402.

(51) Abdelrahman, O. A.; Park, D. S.; Vinter, K. P.; Spanjers, C. S.; Ren, L.; Cho,

H. J.; Vlachos, D. G.; Fan, W.; Tsapatsis, M.; Dauenhauer, P. J. Biomass-

Derived Butadiene by Dehydra-Decyclization of Tetrahydrofuran. ACS Sustain.

Chem. Eng. 2017, 5, 3732–3736.

(52) Somorjai, G. A.; Li, Y. Impact of Surface Chemistry. Proc. Natl. Acad. Sci. U.

S. A. 2011, 108, 917–924.

(53) Kaden, W. E.; Pomp, S.; Sterrer, M.; Freund, H. J. Insights into Silica Bilayer

Hydroxylation and Dissolution. Top. Catal. 2017, 60, 471–480.

(54) Heard, C. J.; Grajciar, L.; Rice, C. M.; Pugh, S. M.; Nachtigall, P.; Ashbrook,

S. E.; Morris, R. E. Fast Room Temperature Lability of Aluminosilicate

Zeolites. Nat. Commun. 2019, 10.

(55) Yang, B.; Emmez, E.; Kaden, W. E.; Yu, X.; Boscoboinik, J. A.; Sterrer, M.;

100

Shaikhutdinov, S.; Freund, H. J. Hydroxylation of Metal-Supported Sheet-like

Silica Films. J. Phys. Chem. C 2013, 117, 8336–8344.

(56) Yu, X.; Emmez, E.; Pan, Q.; Yang, B.; Pomp, S.; Kaden, W. E.; Sterrer, M.;

Shaikhutdinov, S.; Freund, H. J.; Goikoetxea, I.; et al. Electron Stimulated

Hydroxylation of a Metal Supported Silicate Film. Phys. Chem. Chem. Phys.

2016, 18, 3755–3764.

(57) Stockwell, P. B. Handbook of Elemental Speciation: Techniques and

Methodology; 2004.

(58) Kimbrough, R. D. Toxicity and Health Effects of Selected Organotin

Compounds: A Review. Environ. Health Perspect. 1976.

(59) Nakanishi, T. Endocrine Disruption Induced by Organotin Compounds;

Organotins Function as a Powerful Agonist for Nuclear Receptors Rather than

an Aromatase Inhibitor. J. Toxicol. Sci. 2008.

(60) Sunday, A. O.; Alafara, B. A.; Oladele, O. G. Toxicity and Speciation Analysis

of Organotin Compounds. Chem. Speciat. Bioavailab. 2012.

(61) Mesmer, R. E.; Baes, C. F. Review of Hydrolysis Behavior of Ions in Aqueous

Solutions. MRS Proc. 1990, 180, 85.

(62) Jurinak, J. J.; Baes Jr, C. F. Jr., and Mesmer, R. The Hydrolysis of Cations;

1976.

(63) Baes, C. F.; Mesmer, R. F. The Thermodynamics of Cation Hydrolysis. Am. J.

Sci. 1981, 281, 935–962.

(64) Van Berkel, G. J.; Kertesz, V. Using the Electrochemistry of the Electrospray

Ion Source. Analytical Chemistry. 2007.

(65) Urabe, T.; Tanaka, M.; Kumakura, S.; Tsugoshi, T. Study on Chemical

Speciation in Aluminum Chloride Solution by ESI-Q-MS. J. Mass Spectrom.

2007.

(66) Stewart, I. I.; Horlick, G. Investigations into Chromium Speciation by

Electrospray Mass Spectrometry. J. Anal. At. Spectrom. 1996.

(67) Mensah, J. B.; Delidovich, I.; Hausoul, P. J. C.; Weisgerber, L.; Schrader, W.;

Palkovits, R. Mechanistic Studies of the Cu(OH)+-Catalyzed Isomerization of

Glucose into Fructose in Water. ChemSusChem 2018.

(68) Jiang, Y.; Yang, L.; Bohn, C. M.; Li, G.; Han, D.; Mosier, N. S.; Miller, J. T.;

Kenttämaa, H. I.; Abu-Omar, M. M. Speciation and Kinetic Study of Iron

Promoted Sugar Conversion to 5-Hydroxymethylfurfural (HMF) and Levulinic

Acid (LA). Org. Chem. Front. 2015, 2, 1388.

(69) Diehl, B.; Malz, F.; Holzgrabe, U. Quantitative NMR Spectroscopy in the

Quality Evaluation of Active Pharmaceutical Ingredients and Excipients.

Spectrosc. Eur. 2007.

(70) Holzgrabe, U. Quantitative NMR Spectroscopy in Pharmaceutical Applications.

Prog. Nucl. Magn. Reson. Spectrosc. 2010.

(71) Luan, J.; Feng, R.; Yu, C.; Wu, X.; Shen, W.; Chen, Y.; Di, B.; Su, M.

Quantitative Assessment of the Absolute Purity of Thiopeptcin Reference

Standard by 1H-NMR. Anal. Sci. 2018.

101

(72) Maki, H.; Sakata, G.; Mizuhata, M. Quantitative NMR of Quadrupolar Nucleus

as a Novel Analytical Method: Hydrolysis Behaviour Analysis of Aluminum

Ion. Analyst 2017, 142, 1790–1799.

(73) Kalbitzer, H. R. Protein NMR Spectroscopy. Principles and Practice. Zeitschrift

für Phys. Chemie 2011.

(74) Elving, P. J.; Zemel, B. Absorption in the Ultraviolet and Visible Regions of

Chloroaquochromium(III) Ions in Acid Media. J. Am. Chem. Soc. 1957.

(75) Abbott, A. P.; Al-Barzinjy, A. A.; Abbott, P. D.; Frisch, G.; Harris, R. C.;

Hartley, J.; Ryder, K. S. Speciation, Physical and Electrolytic Properties of

Eutectic Mixtures Based on CrCl3·6H2O and Urea. Phys. Chem. Chem. Phys.

2014.

(76) Verdonck, T.; Verpoort, P.; De Strycker, J.; De Cleene, A.; Banerjee, D.;

Nockemann, P.; Van Deun, R.; Van Hecke, K. Combining MCR-ALS and

EXAFS as Tools for Speciation of Highly Chlorinated Chromium(Iii) in

Mixtures of Deep Eutectic Solvents and Water. Dalt. Trans. 2019.

(77) McCalman, D. C.; Sun, L.; Zhang, Y.; Brennecke, J. F.; Maginn, E. J.;

Schneider, W. F. Speciation, Conductivities, Diffusivities, and Electrochemical

Reduction as a Function of Water Content in Mixtures of Hydrated Chromium

Chloride/Choline Chloride. J. Phys. Chem. B 2015.

(78) Yano, J.; Yachandra, V. K. X-Ray Absorption Spectroscopy. Photosynth. Res.

2009.

(79) Apted, M. J.; Waychunas, G. A.; Brown, G. E. Structure and Specification of

Iron Complexes in Aqueous Solutions Determined by X-Ray Absorption

Spectroscopy. Geochim. Cosmochim. Acta 1985.

(80) Stetefeld, J.; McKenna, S. A.; Patel, T. R. Dynamic Light Scattering: A

Practical Guide and Applications in Biomedical Sciences. Biophysical Reviews.

2016.

(81) Chu, B.; Liu, T. Characterization of Nanoparticles by Scattering Techniques. J.

Nanoparticle Res. 2000.

(82) Shafran, K. L.; Perry, C. C. A Systematic Investigation of Aluminium Ion

Speciation at High Temperature. Part. 1. Solution Studies. Dalt. Trans. 2005.

(83) Caetano, B. L.; Meneau, F.; Santilli, C. V.; Pulcinelli, S. H.; Magnani, M.;

Briois, V. Mechanisms of SnO2 Nanoparticles Formation and Growth in Acid

Ethanol Solution Derived from SAXS and Combined Raman-XAS Time-

Resolved Studies. Chem. Mater. 2014.

(84) Kosinski, J. J.; Wang, P.; Springer, R. D.; Anderko, A. Modeling Acid-Base

Equilibria and Phase Behavior in Mixed-Solvent Electrolyte Systems. Fluid

Phase Equilib. 2007.

(85) Wang, P.; Anderko, A.; Springer, R. D.; Young, R. D. Modeling Phase

Equilibria and Speciation in Mixed-Solvent Electrolyte Systems: II. Liquid-

Liquid Equilibria and Properties of Associating Electrolyte Solutions. J. Mol.

Liq. 2006.

(86) Wang, P.; Springer, R. D.; Anderko, A.; Young, R. D. Modeling Phase

102

Equilibria and Speciation in Mixed-Solvent Electrolyte Systems. In Fluid Phase

Equilibria; 2004.

(87) Wang, P.; Anderko, A.; Springer, R. D.; Kosinski, J. J.; Lencka, M. M.

Modeling Chemical and Phase Equilibria in Geochemical Systems Using a

Speciation-Based Model. J. Geochemical Explor. 2010.

(88) Wang, P.; Anderko, A.; Young, R. D. A Speciation-Based Model for Mixed-

Solvent Electrolyte Systems. Fluid Phase Equilib. 2002.

(89) Norton, A. M.; Nguyen, H.; Xiao, N. L.; Vlachos, D. G. Direct Speciation

Methods to Quantify Catalytically Active Species of AlCl 3 in Glucose

Isomerization. RSC Adv. 2018, 8, 17101-17109.

(90) A Guide to Using OLI Analyzer; 2014.

(91) Driscoll, C. T. The Chemistry of Aluminum in Surface Waters. Environ. Chem.

Alum. 1989.

(92) Hem, J. D. Study and Interpretation of the Chemical Characteristics of Natural

Water, 73rd ed.; Gould, R. F., Ed.; American Chemical Society: Washington,

D.C., 1970.

(93) Mesmer, R. E.; Baes, C. F. Acidity Measurements at Elevated Temperatures. V.

Aluminum Ion Hydrolysis. Inorg. Chem. 1971.

(94) Gustafsson, J. P. Visual MINTEQ https://vminteq.lwr.kth.se/ (accessed Jul 16,

2019).

(95) Shaff, J. E.; Schultz, B. A.; Craft, E. J.; Clark, R. T.; Kochian, L. V.

GEOCHEM-EZ: A Chemical Speciation Program with Greater Power and

Flexibility. Plant Soil 2010.

(96) Klamt, A.; Eckert, F.; Arlt, W. COSMO-RS: An Alternative to Simulation for

Calculating Thermodynamic Properties of Liquid Mixtures. Rev. Adv. Annu.

Rev. Chem. Biomol. Eng. 2010 2010, 1, 101–122.

(97) Kiepe, rn; Noll, O.; Gmehling, rgen. Modified LIQUAC and Modified

LIFACsA Further Development of Electrolyte Models for the Reliable

Prediction of Phase Equilibria with Strong Electrolytes. Ind. Eng. Chem. Res.

2006, 45, 2361–2373.

(98) Wang, S.; Song, Y.; Chen, C.-C. Extension of COSMO-SAC Solvation Model

for Electrolytes. Ind. Eng. Chem. Res. 2011, 50, 176–187.

(99) MT, H.; ST, L. A Predictive Model for the Excess Gibbs Free Energy of Fully

Dissociated Electrolyte Solutions. AIChE J. 2011, 57, 1061–1074.

(100) Ingram, T.; Gerlach, T.; Mehling, T.; Smirnova, I. Extension of COSMO-RS

for Monoatomic Electrolytes: Modeling of Liquid-Liquid Equilibria in Presence

of Salts. Fluid Phase Equilib. 2011, 314, 29–37.

(101) Mohammad, S.; Held, C.; Altuntepe, E.; Köse, T.; Gerlach, T.; Smirnova, I.;

Sadowski, G. Salt Influence on MIBK/Water Liquid-Liquid Equilibrium:

Measuring and Modeling with EPC-SAFT and COSMO-RS. Fluid Phase

Equilib. 2016, 416, 83–93.

(102) Mushrif, S. H.; Varghese, J. J.; Vlachos, D. G. Insights into the Cr(Iii)

Catalyzed Isomerization Mechanism of Glucose to Fructose in the Presence of

103

Water Using Ab Initio Molecular Dynamics. Phys. Chem. Chem. Phys. 2014.

(103) Isikgor, F. H.; Remzi Becer, C. Lignocellulosic Biomass: A Sustainable

Platform for Production of Bio-Based Chemicals and Polymers.

(104) Wang, Y.; Deng, W.; Wang, B.; Zhang, Q.; Wan, X.; Tang, Z.; Wang, Y.; Zhu,

C.; Cao, Z.; Wang, G.; et al. Chemical Synthesis of Lactic Acid from Cellulose

Catalysed by Lead(II) Ions in Water. Nat. Commun. 2013, 4.

(105) Tang, Z.; Deng, W.; Wang, Y.; Zhu, E.; Wan, X.; Zhang, Q.; Wang, Y.

Transformation of Cellulose and Its Derived Carbohydrates into Formic and

Lactic Acids Catalyzed by Vanadyl Cations. ChemSusChem 2014, 7, 1557–

1567.

(106) Wang, F. F.; Liu, C. L.; Dong, W. S. Highly Efficient Production of Lactic

Acid from Cellulose Using Lanthanide Triflate Catalysts. Green Chem. 2013,

15, 2091–2095.

(107) Lei, X.; Wang, F. F.; Liu, C. L.; Yang, R. Z.; Dong, W. S. One-Pot Catalytic

Conversion of Carbohydrate Biomass to Lactic Acid Using an ErCl3 Catalyst.

Appl. Catal. A Gen. 2014, 482, 78–83.

(108) Fenn, T. D.; Ringe, D.; Petsko, G. A. Xylose Isomerase in Substrate and

Inhibitor Michaelis States: Atomic Resolution Studies of a Metal-Mediated

Hydride Shift. Biochemistry 2004, 43, 6464–6474.

(109) Delidovich, I.; Palkovits, R. Catalytic Isomerization of Biomass-Derived

Aldoses: A Review. ChemSusChem. 2016.

(110) Rasrendra, C.; Makertihartha, I.; Adisasmito, S.; Heeres, H. Green Chemicals

from D-Glucose: Systematic Studies on Catalytic Effects of Inorganic Salts on

the Chemo-Selectivity and Yield in Aqueous Solutions. Top. Catal. 2010, 53,

1241–1247.

(111) Dallas Swift, T.; Nguyen, H.; Anderko, A.; Nikolakis, V.; Vlachos, D. G.

Tandem Lewis/Brønsted Homogeneous Acid Catalysis: Conversion of Glucose

to 5-Hydoxymethylfurfural in an Aqueous Chromium(Iii) Chloride and

Hydrochloric Acid Solution. Green Chem. 2015.

(112) Pagán-Torres, Y. J.; Wang, T. F.; Gallo, J. M. R.; Shanks, B. H.; Dumesic, J. A.

Production of 5-Hydroxymethylfurfural from Glucose Using a Combination of

Lewis and Brønsted Acid Catalysts in Water in a Biphasic Reactor with an

Alkylphenol Solvent. ACS Catal. 2012, 2, 930–934.

(113) Enslow, K. R.; Bell, A. T. SnCl4-Catalyzed Isomerization/Dehydration of

Xylose and Glucose to Furanics in Water. Catal. Sci. Technol. 2015, No. 5,

2839–2847.

(114) Dos Santos, J. B.; Da Silva, F. L.; Altino, F. M. R. S.; Da Silva Moreira, T.;

Meneghetti, M. R.; Meneghetti, S. M. P. Cellulose Conversion in the Presence

of Catalysts Based on Sn(Iv). Catal. Sci. Technol. 2013, 3, 673–678.

(115) Ramesh, P.; Kritikos, A.; Tsilomelekis, G. Effect of Metal Chlorides on

Glucose Mutarotation and Possible Implications on Humin Formation. React.

Chem. Eng. 2019.

(116) Lee, H. S.; Hong, J. Kinetics of Glucose Isomerization to Fructose by

104

Immobilized Glucose Isomerase: Anomeric Reactivity of D-Glucose in Kinetic

Model. J. Biotechnol. 2000.

(117) Qi, T.; He, M.-F.; Zhu, L.-F.; Lyu, Y.-J.; Yang, H.-Q.; Hu, C.-W. Cooperative

Catalytic Performance of Lewis and Brønsted Acids from AlCl 3 Salt in

Aqueous Solution Toward Glucose-to-Fructose Isomerization. J. Phys. Chem. C

2019.

(118) Corma Canos, A.; Iborra, S.; Velty, A. Chemical Routes for the Transformation

of Biomass into Chemicals. Chem. Rev. 2007, 107, 2411–2502.

(119) Di Cosimo, R.; McAuliffe, J.; Poulose, A. J.; Bohlmann, G. Industrial Use of

Immobilized Enzymes. Chem. Soc. Rev. 2013, 42, 6437–6374.

(120) Bhosale, S. H.; Rao, M. B.; Deshpande, V. V. Molecular and Industrial Aspects

of Glucose Isomerase. Microbiol. Rev. 1996, 60, 280–300.

(121) Nikolla, E.; Rom An-Leshkov, Y.; Moliner, M.; Davis, M. E. “One-Pot”

Synthesis of 5-(Hydroxymethyl)Furfural from Carbohydrates Using Tin-Beta

Zeolite. ACS Catal 2011, 1, 408–410.

(122) Swift, T. D.; Nguyen, H.; Anderko, A.; Nikolakis, V.; Vlachos, D. G. Tandem

Lewis/Brønsted Homogeneous Acid Catalysis: Conversion of Glucose to 5-

Hydoxymethylfurfural in an Aqueous Chromium(Iii) Chloride and

Hydrochloric Acid Solution. Green Chem. 2015, 17, 4725–4735.

(123) Choudhary, V.; Sandler, S. I.; Vlachos, D. G. Conversion of Xylose to Furfural

Using Lewis and Brønsted Acid Catalysts in Aqueous Media. ACS Catal. 2012,

2, 2022–2028.

(124) Bermejo-Deval, R.; Assary, R. S.; Nikolla, E.; Moliner, M.; Román-Leshkov,

Y.; Hwang, S. J.; Palsdottir, A.; Silverman, D.; Lobo, R. F.; Curtiss, L. A.; et al.

Metalloenzyme-like Catalyzed Isomerizations of Sugars by Lewis Acid

Zeolites. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 9727–9732.

(125) Hsu, P. H. Mechanisms of Gibbsite Crystallization from Partially Neutralized

Aluminum Chloride Solutions. Clays Clay Miner. 1988, 36, 25–30.

(126) Sposito, G. The Environmental Chemistry of Aluminum, Second Edition; 1995.

(127) Baes, C. F.; Mesmer, R. E. The Hydrolysis of Cations; 1977.

(128) Tang, J.; Zhu, L.; Fu, X.; Dai, J.; Guo, X.; Hu, C. Insights into the Kinetics and

Reaction Network of Aluminum Chloride-Catalyzed Conversion of Glucose in

NaCl-H2O/THF Biphasic System. ACS Catal. 2017, 7, 256–266.

(129) Bertsch, P. M.; Barnhisel, R. I.; Thomas, G. W.; Layton, W. J.; Smith, S. L.

Quantitative Determination of Aluminum-27 by High-Resolution Nuclear

Magnetic Resonance Spectrometry. Anal. Chem. 1986, 58, 2583–2585.

(130) Curtin, M. A.; Ingalls, L. R.; Campbell, A.; James-Pederson, M. Hydrolysis

Studies and Quantitative Determination of Aluminum Ions Using 27Al NMR.

An Undergraduate Analytical Chemistry Experiment. J. Chem. Educ. 2008, 85,

291–293.

(131) Faust, B. C.; Labiosa, W. B.; Dai, K. H.; MacFall, J. S.; Browne, B. A.;

Ribeiro, A. A.; Richter, D. D. Speciation of Aqueous Mononuclear Al(III)-

Hydroxo and Other Al(III) Complexes at Concentrations of Geochemical

105

Relevance by Aluminum-27 Nuclear Magnetic Resonance Spectroscopy.

Geochim. Cosmochim. Acta 1995, 59, 2651–2661.

(132) Shafran, K.; Deschaume, O.; Perry, C. C. High-Temperature Speciation Studies

of Al-Lon Hydrolysis. Adv. Eng. Mater. 2004, 6, 836–839.

(133) Singhal, A.; Keefer, K. D. A Study of Aluminum Speciation in Aluminum

Chloride Solutions by Small Angle X-Ray Scattering and 27A1 NMR. J. Mater.

Res. 1994, 9, 1973–1983.

(134) Wang, P.; Anderko, A. Computation of Dielectric Constants of Solvent

Mixtures and Electrolyte Solutions. Fluid Phase Equilib. 2001, 186, 103–122.

(135) Wang, P.; Anderko, A.; Young, R. D. A Speciation-Based Model for Mixed-

Solvent Electrolyte Systems. Fluid Phase Equilib. 2002, 203, 123–142.

(136) Baes, C. F.; Mesmer, R. F. The Thermodynamics of Cation Hydrolysis. Am. J.

Sci. 1981, 281, 935–962.

(137) Hem, J. D. Trace Inorganics in Water, 73rd ed.; Gould, R. F., Ed.; American

Chemical Society: Washington, D.C., 1970; Vol. 73.

(138) Mesmer, R. E.; Baes, C. F. Acidity Measurements at Elevated Temperatures. V.

Aluminum Ion Hydrolysis. Inorg. Chem. 1971, 10, 2290–2296.

(139) Frink, C. R.; Peech, M. Hydrolysis of the Aluminum Ion in Dilute Aqueous

Solutions. Inorg. Chem. 1963, 2, 473–478.

(140) Mason, J. Multinuclear NMR. In Multinuclear NMR; 1987; pp 259–292.

(141) Malz, F.; Jancke, H. Validation of Quantitative NMR. J. Pharm. Biomed. Anal.

2005, 38, 813–823.

(142) Pauli, G. F.; Jaki, B. U.; Lankin, D. C. Quantitative 1H NMR: Development

and Potential of a Method for Natural Products Analysis. J. Nat. Prod. 2005,

68, 133–149.

(143) Lubricants Market Size, Share & Trends Analysis Report By Application

(Industrial, Automotive, Marine, Aerospace), By Region (North America,

Europe, APAC, CSA, MEA), Competitive Landscape, And Segment Forecast,

2018-2025 https://www.grandviewresearch.com/industry-analysis/lubricants-

market.

(144) Liu, S.; Dutta, S.; Zheng, W.; Gould, N. S.; Cheng, Z.; Xu, B.; Saha, B.;

Vlachos, D. G. Catalytic Hydrodeoxygenation of High Carbon Furylmethanes

to Renewable Jet-Fuel Ranged Alkanes over a Rhenium-Modified Iridium

Catalyst. ChemSusChem 2017, 10, 3225–3234.

(145) Raston, C. L.; Scott, J. L. Chemoselective, Solvent-Free Aldol Condensation

Reaction. Green Chem. 2000, 2, 49–52.

(146) Del Rio, J. C.; Philp, R. P. Oligomerization of Fatty Acids as a Possible Source

for High Molecular Weight Hydrocarbons and Sulphur-Containing Compounds

in Sediments. Org. Geochem. 1992, 18, 869–880.

(147) Wijaya, Y. P.; Kristianto, I.; Lee, H.; Jae, J. Production of Renewable Toluene

from Biomass-Derived Furans via Diels-Alder and Dehydration Reactions: A

Comparative Study of Lewis Acid Catalysts. Fuel 2016, 182, 588–596.

(148) Corma, A. From Microporous to Mesoporous Molecular Sieve Materials and

106

Their Use in Catalysis. Chem. Rev. 1997, 97, 2373–2420.

(149) Davis, M. E. Ordered Porous Materials for Emerging Applications. Nature

2002, 417, 813–821.

(150) Meininghaus, C. K. W.; Prins, R. Sorption of Volatile Organic Compounds on

Hydrophobic Zeolites. Microporous Mesoporous Mater. 2000, 35, 349–365.

(151) Kreiter, R.; Rietkerk, M. D. A.; Castricum, H. L.; Van Veen, H. M.; Ten Elshof,

J. E.; Vente, J. F. Evaluation of Hybrid Silica Sols for Stable Microporous

Membranes Using High-Throughput Screening. J. Sol-Gel Sci. Technol. 2011,

57, 242–252.

(152) Luo, G.; McDonald, A. G. Conversion of Methanol and Glycerol into Gasoline

via ZSM-5 Catalysis. Energy and Fuels 2014, 28, 600–606.

(153) Kim, D.; Jeon, M. Y.; Stottrup, B. L.; Tsapatsis, M. Para-Xylene Ultra-

Selective Zeolite MFI Membranes Fabricated from Nanosheet Monolayers at

the Air–Water Interface. Angew. Chemie - Int. Ed. 2018, 57, 480–485.

(154) Kim, D.; Shete, M.; Tsapatsis, M. Large-Grain, Oriented, and Thin Zeolite MFI

Films from Directly Synthesized Nanosheet Coatings. Chem. Mater. 2018, 30,

3545–3551.

(155) Flanigen, E. M.; Khatami, H.; Szymanski, H. A. Infrared Structural Studies of

Zeolite Frameworks; 1974.

(156) Kestell, J. D.; Zhong, J. Q.; Shete, M.; Waluyo, I.; Sadowski, J. T.; Stacchiola,

D. J.; Tsapatsis, M.; Boscoboinik, J. A. Studying Two-Dimensional Zeolites

with the Tools of Surface Science: MFI Nanosheets on Au(111). Catal. Today

2017, 280, 283–288.

(157) Bisio, C.; Martra, G.; Coluccia, S.; Massiani, P. FT-IR Evidence of Two

Distinct Protonic Sites in BEA Zeolite: Consequences on Cationic Exchange

and on Acido-Basic Properties in the Presence of Cesium. J. Phys. Chem. C

2008, 112, 10520–10530.

(158) Warring, S. L.; Beattie, D. A.; McQuillan, A. J. Surficial Siloxane-to-Silanol

Interconversion during Room-Temperature Hydration/Dehydration of

Amorphous Silica Films Observed by ATR-IR and TIR-Raman Spectroscopy.

Langmuir 2016, 32, 1568–1576.

(159) Karbowiak, T.; Saada, M. A.; Rigolet, S.; Ballandras, A.; Weber, G.;

Bezverkhyy, I.; Soulard, M.; Patarin, J.; Bellat, J. P. New Insights in the

Formation of Silanol Defects in Silicalite-1 by Water Intrusion under High

Pressure. Phys. Chem. Chem. Phys. 2010, 12, 11454–11466.

(160) Maier, S. M.; Jentys, A.; Lercher, J. A. Steaming of Zeolite BEA and Its Effect

on Acidity: A Comparative NMR and IR Spectroscopic Study. J. Phys. Chem.

C 2011, 115, 8005–8013.

(161) Zhuravlev, L. T. The Surface Chemistry of Amorphous Silica. Zhuravlev

Model. Colloids Surfaces A Physicochem. Eng. Asp. 2000, 173, 1–38.

(162) Gallas, J. P.; Goupil, J. M.; Vimont, A.; Lavalley, J. C.; Gil, B.; Gilson, J. P.;

Miserque, O. Quantification of Water and Silanol Species on Various Silicas by

Coupling IR Spectroscopy and In-Situ Thermogravimetry. Langmuir 2009, 25,

107

5825–5834.

(163) Yang, Y.; Xia, Q.; Feng, M.; Zhang, P. Temperature Dependence of IR

Absorption of OH Species in Clinopyroxene. Am. Mineral. 2010, 95, 1439–

1443.

(164) Morrow, B. A.; Devi, A. Adsorption of Water and 18 O Exchange of Surface

Hydroxyl Groups on Silica. Can. J. Chem. 1970, 48, 2454–2456.

(165) Morrow, B. A.; Cody, I. A.; Lee, L. S. M. Infrared Studies of Reactions on

Oxide Surfaces. 7. Mechanism of the Adsorption of Water and Ammonia on

Dehydroxylated Silica. J. Phys. Chem. 1976, 80, 2761–2767.

(166) Hoffmann, P.; Knözinger, E. Novel Aspects of Mid and Far IR Fourier

Spectroscopy Applied to Surface and Adsorption Studies on SiO2. Surf. Sci.

1987, 188, 181–198.

(167) Sun, S. T.; Jiang, L.; Liu, J. W.; Heine, N.; Yacovitch, T. I.; Wende, T.; Asmis,

K. R.; Neumark, D. M.; Liu, Z. F. Microhydrated Dihydrogen Phosphate

Clusters Probed by Gas Phase Vibrational Spectroscopy and First Principles

Calculations. Phys. Chem. Chem. Phys. 2015, 17, 25714–25724.

(168) Allcock, H. R. Introduction to Materials Chemistry, Second.; John Wiley &

Sons, Ltd: Hoboken, 2008.

(169) Rumplmayr, G.; Lercher, J. A. Modification of HZSM-5 with

Trimethylphosphine. Zeolites 1990, 10, 283–287.

(170) Kycia, A. H.; Vezvaie, M.; Zamlynny, V.; Lipkowski, J.; Petryk, M. W. P.

Non-Contact Detection of Chemical Warfare Simulant Triethyl Phosphate

Using PM-IRRAS. Anal. Chim. Acta 2012, 737, 45–54.

(171) Hadjiivanov, K. Identification and Characterization of Surface Hydroxyl

Groups by Infrared Spectroscopy. In Advances in Catalysis; 2014; pp 99–313.

108

SUPPLEMENTARY INFORMATION FOR CHAPTER 2

A.1 SnCl4 Speciation (5 mM) in Water at 413 K

Appendix A

Figure A-1: (a) Distribution of hydrolyzed SnCl4 species (5 mM SnCl4) at 413 K,

calculated using the OLI speciation model. The hydration sphere of water has been

removed for clarity. (b) Zoomed in profile of the dominate SnO2(solid) species.

dominate SnO2(solid) species.

0 20 40 60 80 100 120 1400.00E+00

2.00E-11

4.00E-11

6.00E-11

8.00E-11

Sn4+

SnOH3+

Sn(OH)2+2

Sn(OH)1-5

Sn

(OH

) 4,

Sn

(OH

)1+

3 (

mM

)

Sn

4+,

Sn

OH

3+,

Sn

(OH

)2+

2,

Sn

(OH

)1-

5 (

mM

)

HCl (mM)

0.00

2.00x10-5

4.00x10-5

6.00x10-5

8.00x10-5

1.00x10-4

1.20x10-4

1.40x10-4

0 20 40 60 80 100 120 1404.99976

4.99978

4.99980

4.99982

4.99984

Sn

O2

(so

lid

) (m

M)

HCl (mM)

SnO2 (solid)

Sn(OH)4

Sn(OH)1+3

0 20 40 60 80 100 120 1404.99976

4.99978

4.99980

4.99982

4.99984

Sn

O2

(so

lid

) (m

M)

HCl (mM)

SnO2 (solid)

(a)

(b)

109

A.2 CrCl3 Speciation (5 mM) in Water at 413 K

Figure A-2: Distribution of CrCl3 speciation (5 mM CrCl3) at 413 K, calculated using

the OLI speciation model. The hydration sphere of water has been removed for clarity.

0 20 40 60 80 1000

1

2

3

CrO

1+ (

mM

)

CrC

l2+, C

r(O

H) 3

(so

lid

), C

rOH

2+ (

mM

)

HCl (mM)

CrCl2+

Cr(OH)3(solid)

CrOH2+

0.00

1.00x10-2

2.00x10-2

3.00x10-2

4.00x10-2

5.00x10-2

6.00x10-2

CrO1+

110

A.3 SnCl4 Species at Varying Concentrations


(a) SnO2(solid), (b) SnO2(solid) zoomed in profile, (c) Sn(OH)4, (d) Sn(OH)31+, (e)

Sn(OH)22+, and (f) Sn(OH)5

1- as a function of HCl concentration, where SnCl4

concentration is 5, 15, and 30 mM. zoomed in profile.

0 20 40 60 80 1000

4

8

12

16

20

24

28

Sn

O2

(so

lid

) (m

M)

HCl (mM)

5 mM SnCl4

15 mM SnCl4

30 mM SnCl4

0 20 40 60 80 1000.0001348

0.0001352

0.0001356

0.0001360

0.0001364

Sn

(OH

) 4 (m

M)

HCl (mM)

5 mM SnCl4

15 mM SnCl4

30 mM SnCl4

0 20 40 60 80 1000.00000

0.00002

0.00004

0.00006

0.00008

0.00010

Sn

(OH

)1+

3 (m

M)

HCl (mM)

5 mM SnCl4

15 mM SnCl4

30 mM SnCl4

0 20 40 60 80 1000.00

2.00E-11

4.00E-11

6.00E-11

Sn

(OH

)2+

2 (m

M)

HCl (mM)

5 mM SnCl4

15 mM SnCl4

30 mM SnCl4

0 20 40 60 80 1000.00

2.00E-11

4.00E-11

6.00E-11

Sn

(OH

)1-

5 (

mM

)

HCl (mM)

5 mM SnCl4

15 mM SnCl4

30 mM SnCl4

0 20 40 60 80 1004.99979

4.99980

4.99981

4.99982

4.99983

4.99984

4.99985

Sn

O2

(so

lid

) (m

M)

HCl (mM)

5 mM SnCl4

(c) (d)

(a) (b)

(c) (d)

HCl (mM)

HCl (mM)

HCl (mM)

HCl (mM)

HCl (mM) HCl (mM)

111

A.4 AlCl3 Species at Varying Concentrations


(a) Al(OH)21+, (b) Al(OH)2+, (c) AlO(OH)solid, and (d) Al(OH)3 as a function of HCl

concentration, where AlCl3 concentration is 5, 15, and 30 mM.

0 20 40 60 80 1000.000

0.002

0.004

0.006

0.008

0.010

0.012

0.014

Al(

OH

)1+

2 (

mM

)

HCl (mM)

5 mM AlCl3

15 mM AlCl3

30 mM AlCl3

0 20 40 60 80 100

0

2

4

6

8

AlO

(OH

) so

lid (

mM

)

HCl (mM)

5 mM AlCl3

15 mM AlCl3

30 mM AlCl3

0 20 40 60 80 1000.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

Al(

OH

)2+ (

mM

)HCl (mM)

5 mM AlCl3

15 mM AlCl3

30 mM AlCl3

0 20 40 60 80 100

0.00000

0.00005

0.00010

0.00015

0.00020

Al(

OH

) 3 (

mM

)

HCl (mM)

5 mM AlCl3

15 mM AlCl3

30 mM AlCl3

(a) (b)

(c) (d)

HCl (mM)

HCl (mM)HCl (mM)

HCl (mM)

112

A.5 CrCl3 Species at Varying Concentrations


(a) CrOH2+, (b) CrO1+, (c) CrCl2+, and (d) Cr(OH)3(solid) as a function of HCl

concentration, where CrCl3 concentration is 5, 15, and 30 mM.

0 20 40 60 80 1000

1

2

3

4

5

Cr(

OH

)2+ (

mM

)

HCl (mM)

5 mM CrCl3

15 mM CrCl3

30 mM CrCl3

0 20 40 60 80 100

0

1

2

3

Cr(

OH

) 3 (

soli

d) (m

M)

HCl (mM)

5 mM CrCl3

15 mM CrCl3

30 mM CrCl3

0 20 40 60 80 1000

1

2

3

4

5

CrC

l2+ (

mM

)

HCl (mM)

5 mM CrCl3

15 mM CrCl3

30 mM CrCl3

0 20 40 60 80 100

0.00

0.02

0.04

0.06

0.08

CrO

1+ (

mM

)

HCl (mM)

5 mM CrCl3

15 mM CrCl3

30 mM CrCl3

(a) (b)

(c) (d)

HCl (mM)

HCl (mM)

HCl (mM)

HCl (mM)

113


B.1 Assessment of 3300 HT pH probe and qNMR

Appendix B

1.65

1.7

1.75

1.8

1.85

10 35 60 85 110

pH

Temperature (oC)

(a)

2.99

3

3.01

3.02

3.03

3.04

3.05

15 25 35 45 55

pH

Temperature (oC)

(b)

Figure B-1: pH measurements obtained on the 3300 HT pH probe (circles) as a

function of temperature for the (a) pH = 1.678 and (b) pH = 3 buffer solutions. The

solid lines represent the expected pH. Solution temperature control was applied.

114

0

1

2

3

4

5

6

0 50 100 150 200

Al3

+(m

M)

HCl (mM)

Figure B-2: Concentration of Al3+ measured by 27Al qNMR (closed circles) as a

function of HCl concentration, along with OLI model predictions (line). Conditions:

303 K, 5 mM AlCl3 with HCl concentration between 0 and 200 mM.

115

B.2 Aluminum Speciation as Function of Temperature

0

1

2

3

4

5

6

0 20 40 60 80 100

Al3

+ (

mM

)

HCl (mM)

30 C

90 C

140 C

0

0.02

0.04

0.06

0.08

0 20 40 60 80 100

Al(

OH

)2+

(mM

)

HCl (mM)

30 C

90 C

140 C

0

0.002

0.004

0.006

0.008

0.01

0.012

0 20 40 60 80 100

Al(

OH

) 21

+ (m

M)

HCl (mM)

30 C

90 C

140 C

303 K

363 K

413 K

0

1

2

3

4

5

0 20 40 60 80 100

AlO

(OH

) (m

M)

HCl (mM)

30 C

90 C

140 C

303 K

363 K

413 K

303 K

363 K

413 K

303 K

363 K

413 K

(a) (b)

(c) (d)

Figure B-3: Effect of temperature on the distribution of (a) Al3+, (b) AlO(OH), (c)

Al(OH)2+, and (d) Al(OH)21+ in 5 mM AlCl3 as a function of HCl concentration,

calculated using the OLI software.

116

B.3 Ex Situ and In Situ pH Measurements

0

0.5

1

1.5

2

2.5

3

0 10 20 30 40 50

pH

HCl (mM)

24 hr

Series3

pH

Ex situ, pre-heat, 413 K

In situ, 413 K


1

1.5

2

2.5

3

3.5

4

4.5

0 10 20 30 40 50

pH

HCl (mM)

30 C

Speciation Model, 30 C

In situ (303 K)


(a) (b)

Figure B-4: pH measured ex situ (closed circles) or in situ (open circles) compared to

pH calculated using the OLI speciation software (lines) at (a) 303 K and (b) 413 K.

117

B.4 Equilibration of AlCl3-HCl Catalyst Solutions

0

40

80

120

160

200

0 8 16 24 32 40 48

Pa

rtic

le D

iam

eter

(n

m)

Time (h)

0 mM HCl

3 mM HCl

10 mM HCl

0

40

80

120

160

200

0 8 16 24 32 40 48

Pa

rtic

le D

iam

eter

(n

m)

Time (h)

0 mM HCl

3 mM HCl

10 mM HCl

(a) (b)

Figure B-5: Average particle diameter (nm), obtained from DLS, as a function of time.

Experimental conditions: 5 mM AlCl3 and HCl (specified), preheated at (a) 413 K and

(b) 363 K for 24 h and cooled to room temperature.

1

1.5

2

2.5

3

3.5

4

0 8 16 24 32 40 48

pH

Time (h)

0 mM HCl

3 mM HCl

10 mM HCl

20 mM HCl

44 mM HCl

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 8 16 24 32 40 48

pH

Time (h)

0 mM HCl3 mM HCl10 mM HCl20 mM HCl44 mM HCl

(a) (b)

Figure B-6: (a) pH and (b) amount of solid vs. time at different HCl initial

concentrations. Samples contained 5 mM AlCl3 and were cooled from 363 K to 303 K

prior to measurements.

118

B.5 27Al qNMR Quantification

Table B-1: Data of 27Al qNMR spectra, obtained at 363 K for 24 h, 5 mM AlCl3.

HCl

Chemical

Shift

(ppm)

Chemical

Shift

(Hz)

Intensity

(a.u.)

Width

(Hz)

Normalized

Area

0 0.78 ± 0.02 82.6 ± 0.25 403.5 ± 0.50 7.25 ± 0.05 0.16 ± 0.04

3 0.81 ± 0.02 84.4 ± 0.25 510.1 ± 0.51 7.14 ± 0.03 0.20 ± 0.05

10 0.81 ± 0.05 84.3 ± 0.50 799.9 ± 0.75 7.10 ± 0.09 0.45 ± 0.05

20 0.81 ± 0.02 84.0 ± 0.25 1072.5 ± 0.25 6.76 ± 0.09 0.93 ± 0.06

44 0.81 ± 0.07 84.2 ± 0.74 1232.4 ± 0.25 6.84 ± 0.08 1.00 ± 0.05

100 0.79 ± 0.07 82.7 ± 0.74 1249.8 ± 0.86 6.84 ± 0.06 1.00 ± 0.11

δ (ppm)

(f) 100 mM HCl

external standard

(e) 44 mM HCl

(d) 20 mM HCl

(c) 10 mM HCl

(b) 3 mM HCl

(a) 0 mM HCl

Figure B-7: 27Al qNMR spectra for Al3+. Experimental conditions: preheat 363 K for

24 h, 5 mM AlCl3. qNMR measurements obtained at 363 K.

119

Table B-2: Data of 27Al qNMR spectra, obtained at 413 K for 24 h, 5 mM AlCl3.

The 27Al qNMR spectra were processed using the Mestrelab Research software

(mNOVA). Apodization was 5.09 Hz, zero filling was 64 K, spectra underwent

manual phasing, and automatic baseline correction was applied. Error margins

correspond to 95% confidence level.

HCl Chemical

Shift

(ppm)

Chemical

Shift

(Hz)

Intensity

(a.u.) Width

(Hz) Normalized

Area

0 1.81 ± 0.08 188.0 ± 2.52 360.03 ± 4.93 7.59 ± 0.01 0.73 ± 0.05

3 1.78 ± 0.14 185.6 ± 2.69 378.6 ± 0.96 7.58 ± 0.01 0.82 ± 0.16

10 1.80 ± 0.09 187.7 ± 2.37 835.4 ± 0.89 7.55 ± 0.01 0.99 ± 0.13

20 1.82 ± 0.02 190.2 ± 2.42 847.2 ± 0.47 7.51 ± 0.01 1.00 ± 0.14

44 1.81 ± 0.02 188.2 ± 0.30 846.2 ± 1.19 7.48 ± 0.01 1.00 ± 0.08

100 1.79 ± 0.01 186.7 ± 0.56 848.6 ± 1.74 7.50 ± 0.04 1.00 ± 0.00

120

B.6 Dynamic Light Scattering (DLS)

0.9

1

1.1

1.2

1.3

1.4

1.5

1.6

0.1 100 100000

Inte

nsi

ty A

uto

corr

ela

tio

n

Time (μs)

Pre-filter

Post-filter

Figure B-8: DLS spectra of pre-filtered and post-filtered catalyst solution.

Experimental conditions: 5 mM AlCl3, 0 mM HCl, preheat at 413 K (24 h) and cooled

to room temperature.

121

B.7 Measured Aluminum Speciation


from experiments.


from experiments.

The equilibrium constants obtained from the OLI software were:

(413 K)

KA,1 = 0.28

KA,2 = 0.91

(363 K)

KA,1 = 0.024

KA,2 = 0.013

[Al(H2O)6]3+ + H2O ↔ [Al(H2O)5(OH)]2+ + H3O

+ KA,1

[Al(H2O)6]3+ + 2 H2O ↔ [Al(H2O)4(OH)2]

2+ + 2 H3O+ KA,2

HCl (mM) Al3+ (mM) Solid Al (mM) Al(OH)2+ (mM) Al(OH)21+ (mM)

0 3.67 ± 0.19 1.04 ± 0.19 0.027 ± 0.002 0.004 ± 0.001

3 4.02 ± 0.19 0.36 ± 0.19 0.029 ± 0.006 0.005 ± 0.001

10 4.79 ± 0.19 0.03 ± 0.19 0.014 ± 0.002 0.001 ± 0.000

20 4.70 ± 0.19 0 ± 0.19 0.007 ± 0.001 0.001 ± 0.000

44 4.87 ± 0.19 0 ± 0.19 0.003 ± 0.000 0.000 ± 0.000

HCl (mM) Al3+ (mM) Solid Al (mM) Al(OH)2+ (mM) Al(OH)21+ (mM)

0 0.80 ± 0.19 3.69 ± 0.08 0.024 ± 0.006 0.009 ± 0.002

3 1.00 ± 0.24 3.50 ± 0.03 0.029 ± 0.007 0.010 ± 0.002

10 2.25 ± 0.24 2.54 ± 0.03 0.057 ± 0.006 0.017 ± 0.001

20 4.65 ± 0.32 0.00 ± 0.00 0.061 ± 0.004 0.009 ± 0.001

44 4.99 ± 0.24 0.00 ± 0.00 0.040 ± 0.002 0.004 ± 0.001

122

B.8 Initial Rate Constant at 413 K

0

0.05

0.1

0.15

0.2

0.25

0 10 20 30

-ln

(cg

luco

se/c

glu

cose

, 0)

Time (min)

0 mM HCl

3 mM HCl

10 mM HCl

20 mM HCl

44 mM HCl

0

20

40

60

80

100

120

0

2

4

6

8

10

12

14

0 5 10 15 20 25 30

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

Glucose

Fructose

Mannose

Carbon Balance

0

20

40

60

80

100

120

0

2

4

6

8

10

12

14

0 5 10 15 20 25 30

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

0

20

40

60

80

100

120

0

2

4

6

8

10

12

14

0 5 10 15 20

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

0

20

40

60

80

100

120

0

2

4

6

8

10

12

14

0 5 10 15 20 25 30

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

0

20

40

60

80

100

120

0

2

4

6

8

10

12

14

0 10 20 30

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

3 mM HCl 10 mM HCl

20 mM HCl 44 mM HCl

(c) (d)

(e) (f)

Figure B-9: (a) –ln(cglucose/cglucose,0) vs. time, (b) – (f) conversion of glucose and carbon

balance vs. time. Catalyst solutions were preheated at 413 K for 24 h prior to kinetic

study. Reaction conditions: glucose 1 wt %, Al to glucose molar ratio 9 : 100, 413 K, ,

and either (b) 0 mM, (c) 3 mM, (d) 10 mM, (e) 20 mM, or (f) 44 mM HCl.

123

B.9 Initial Rate Constant at 363 K

0

0.02

0.04

0.06

0.08

0 30 60 90 120 150 180

-ln

(cg

luco

se/c

glu

cose

,0)

Time (min)

0 mM HCl

3 mM HCl

10 mM HCl

20 mM HCl

44 mM HCl

0

20

40

60

80

100

120

0

2

4

6

8

0 30 60 90 120 150 180

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

Glucose

Fructose

Mannose

Carbon Balance

0 mM HCl

0

20

40

60

80

100

120

0

2

4

6

8

0 30 60 90 120 150 180

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

3 mM HCl

0

20

40

60

80

100

120

0

1

2

3

4

5

0 30 60 90 120 150 180

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

10 mM HCl

0

20

40

60

80

100

120

0

1

2

3

4

5

0 30 60 90 120 150 180

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

20 mM HCl

0

20

40

60

80

100

120

0

1

2

3

4

5

0 30 60 90 120 150 180

Ca

rbo

n B

ala

nce

(%

)

Co

nv

ersi

on

(G

luco

se)

or

Yie

ld (

%)

Time (min)

44 mM HCl

(a)(b)

(c) (d)

(e) (f)

Figure B-10: (a) –ln(cglucose/cglucose,0) vs. time, (b) – (f) conversion of glucose and

carbon balance vs. time. Catalyst solutions were preheated at 363 K for 24 h prior to

kinetic study. Reaction conditions: glucose 1 wt %, Al to glucose molar ratio 9 : 100,

363 K, and either (b) 0 mM, (c) 3 mM, (d) 10 mM, (e) 20 mM, or (f) 44 mM HCl.

124

B.10 Identifying the Active Species at 363 K

0

0.0001

0.0002

0.0003

0.0004

0 0.25 0.5 0.75 1

r glu

cose

(min

-1)

Al3+ / Al3+(max)

0

0.0001

0.0002

0.0003

0.0004

0 0.25 0.5 0.75 1

r glu

cose

(min

-1)

Solid Al / Solid Al (max)

0

0.0001

0.0002

0.0003

0.0004

0 0.25 0.5 0.75 1

r glu

cose

(min

-1)

Al(OH)2+ / Al(OH)2+ (max)

0

0.0001

0.0002

0.0003

0.0004

0 0.25 0.5 0.75 1

r glu

cose

(min

-1)

Al(OH)21+ / Al(OH)2

1+(max)

(a) (b)

(c) (d)

Figure B-11: Glucose conversion as a function of measured Al species’

concentrations, normalized to the maximum observed species’ concentration (see

Table B-4 for observed species’ concentrations). Catalyst solutions were preheated at

363 K for 24 h prior to kinetic study. Reaction conditions: glucose 1 wt %, Al to

glucose molar ratio 9 : 100, 363 K.

125


C.1 Micro-Viscometer Apparatus and Standard Measurements

Table C-1: Viscosities measured by the micro-viscometer compared to the reported

viscosities.

Appendix C

Base Oil KV100 (cSt) KV40 (cSt)

Measured Reported Measured Reported

Cannon ® N35 Standard 6.17 5.56 36.04 32.82

Exxon Mobil SpectraSyn PlusTM PAO-4

(group IV) 4.15 4.10 17.13 18.40

Apply pressure

(a) (b)

Figure C-1: (a) Illustration and (b) photo of micro-viscometer (Cannon, calibrated

model #: 9722-H62) used to obtain kinematic viscosities (KVs) in accordance with

ASTM D445.

126

C.2 Proposed C-C Coupling Mechanism in C33 Alkane

C5

C10 C13

C5

C15C4 C4

(a)

(b)

2

2

Figure C-2: (a) Carbon-carbon (C-C) cracking in the tertiary carbon positions of the

C33 alkane to produce alkane byproducts. (b) C-C cracking in secondary carbon

positions; the dashed blue line is another possible route to obtain the C14 alkane

instead of the C15 alkane.

127

C.3 GC Chromatographs for Aldol Condensation and HDO Products

0 2 4 6 8 10 12 14 16 18 20 22

Inte

nsi

ty

Retention time (min)

C33 furan (cis/trans isomers)

C28 furan (cis/trans isomers)

(a) AC

(b) HDO

R = C10H21

C13 alkane

C14 alkane

C15 alkane

C28 alkane (chiral isomers)

C33 alkane (chiral isomers)

Solvent, Dichloromethane

Solvent, Cyclohexane

m/z: 72 – 86

C5, C6 alkanes

C10 alkane

0 2 4 6 8 10 12 14 16 18 20 22

Inte

nsi

ty

Retention time (min)

Figure C-3: (a) GC spectra of aldol condensation (AC) product obtained from a

reaction of furfural (0.45 g) and 12-tricosanone (0.10 g). Potential isomers are

highlighted in the panels on the right. (b) GC spectra of product obtained after HDO of

aldol condensation product. AC products are C28 and C33 furan isomers (0.40 g). HDO

products are primarily chiral isomers of the C28 and C33 alkanes. Eicosane (0.10 g) is

used the internal standard to quantify products, post-reaction, in (a) and (b). m/z was

obtained separately, using GC-MS.

128

C.4 HRMS-LIFDI Chromatographs for Aldol Condensation and HDO Products

*

*

C33H50O3

m/z: 494.75

C28H48O2

m/z: 416.68

C33H68

m/z: 464.89

C28H58

m/z: 394.76

(a) AC

(b) HDO*

*

Figure C-4: HRMS-LIFDI to detect the mass fragments of products after (a) aldol

condensation and (b) hydrodeoxygenation reactions. Solvents for both reactions were

removed prior to analysis, and samples were prepared in dichloromethane (1 mg/mL).

m/z was obtained from HRMS-LIFDI.

129

C.5 NMR Results for Aldol Condensation and HDO Products

C33 furan (predominant product) + isomers

1 2 3 4

76

5

8

1

2

3456

7

8

Chemical Shift (ppm)

8 7 6 5 4 3 2 1 0

C28 furan (secondary product) + isomers

C28H48O2C33H50O3

Figure C-5: 1H NMR spectrum to characterize aldol condensation product. Sample

was prepared in CDCl3 (1 mg/mL). The predominant product is the C33 furan,

followed by the C28 furan, and their isomers. The highlighted bonds in C33H50O3 are

referencing the hydrogen atoms.

130

1 2 6 9

3 5

7C33 furan (predominant product) + isomers C28 furan (secondary product) + isomers

C28H48O2C33H50O3

12

37

4

6

5

8

9

Figure C-6: 13C NMR spectrum to characterize aldol condensation product. Sample

was prepared in CDCl3 (1 mg/mL). The highlighted bonds in C33H50O3 are referencing

the carbon atoms.

131

1 2

1

2

C28H58

C28 alkane (secondary product) + isomersC33 alkane (predominant product) + isomers

C33H68

Figure C-7: 1H NMR spectrum to characterize hydrodeoxygenation product. Sample

was prepared in CDCl3 (1 mg/mL). The highlighted bonds in C33H68 are referencing

the hydrogen atoms. High molecular weight oxygenates are likely present, as indicated

by the chemical shifts at ~3.7 ppm.

132

1 2

1

2

C28H58

C28 alkane (secondary product) + isomers

C28H58

C33 alkane (secondary product) + isomers

Figure C-8: 13C NMR spectrum to characterize hydrodeoxygenation product. Sample

was prepared in CDCl3 (1 mg/mL). The highlighted bonds in C33H68 are referencing

the carbon atoms. Chemical shifts above 60 ppm may correspond to the carbons

associated with the unidentified oxygenates.

133


D.1 SEM and AFM Images of MFI Nanosheets

Appendix D

2 nm-50

-25

0

25

50

5 µm

(a)

(b)

Figure D-1: (a) SEM and (b) AFM images of MFI nanosheets on Au(111) taken in air.

In (a), the MFI nanosheets almost entirely cover the surface (coverage ~1), with some

overlapping regions between corners of some of the nanosheets (which appear darker

gray in the image). The color scale on the right-hand side of (b) shows the height

differences (in nm) on the surface. While most of each nanosheet surface is flat, with a

thickness of ~7 nm, they have a thicker region in the middle corresponding to the seed

material used for their growth. Note the SEM and AFM images were taken on two

separate regions of the material.

134

D.2 XPS of MFI Nanosheets and Bilayer Silicate

-202468101214

Inte

nsi

ty (

a.u

.)

Binding Energy Rel to EF (eV)

82.5083.7585.0086.25

Inte

nsi

ty (

a.u

.)


100101102103104105106107108

Inte

nsi

ty (

a.u

.)


530531532533534535536537

Inte

nsi

ty (

a.u

.)


O 1s533.4

(a)

(c)

(b)

Au 4f

(d)

VB

Si 2p104.1

83.9

Figure D-2: The XPS of MFI on Au(111) was taken in ultra-high vacuum (UHV) at 300 K,

prior to hydrothermal treatment.

135

-20246810

Inte

nsi

ty (

a.u

.)

Binding Energy Relative to EF (eV)

100101102103104105106107In

ten

sity

(a.u

.)

Binding Energy Relative to EF (eV)

528.0529.5531.0532.5534.0535.5537.0

Inte

nsi

ty (

a.u

.)


276280284288

Inte

nsi

ty (

a.u

.)


Si 2p 103.5

VB

(a) (b)

Ru 3d

(d)(c)

O 1s532.8

284.15

279.95

Figure D-3: The XPS of bilayer silicate was taken in UHV at 300 K, prior to

hydrothermal treatment.

136

D.3 Additional IRRAS of MFI Nanosheets and Bilayer Silicate

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

Ra

tio

of

SiO

H:S

iOH

30

0K

, U

HV p

eak

are

as

35003600370038003900

Ab

sorb

an

ce

Wavenumber (cm-1)

3724 cm-1

300 K, UHV

573 K, UHV

573 K, 3 mbar

300 K, UHV

(a) (b)

Pre-H2O

Post-H2O

0.20%

Figure D-4: (a) IRRAS of MFI nanosheets supported on Au(111). The effect of H2O

on SiOH formation in MFI nanosheets supported on Au(111). Pre-, post-, 573 K under

UHV, and 573 K under 3 mbar H2O for 1 h are specified. (b) Quantification of the

IRRAS peak areas for SiOH, taken relative to the initial amount of SiOH obtained

prior to hydrothermal conditions at 300 K and UHV conditions. The peak areas were

obtained from the spectra plotted in absorbance; not much difference was observed

between absorbance and transmission spectra.

137

10001100120013001400R

efle

ctan

ceWavenumber (cm-1)

10001100120013001400

Ref

lect

an

ce

Wavenumber (cm-1)

573 K, 3 mbar H2O

Pre-H2O

Post-H2O

Pre-H2O 573 K, 3 mbar H2O

Post-H2O

(a) (b)

5% 0.5%

Figure D-5: The effect of H2O on the crystalline structure of (a) MFI nanosheets

supported on Au(111) and (b) a polymorphous bilayer silicate supported on Ru(0001).

Pre-, post-, and during (573 K, 3 mbar) H2O treatment are labeled.

138

2600270035003600370038003900R

efle

cta

nce

Wavenumber (cm-1)

3736 cm-12760 cm-1

2600270035003600370038003900

Ref

lect

an

ce

Wavenumber (cm-1)

3724 cm-1 2740 cm-1

Si-OH

Si-OD0.20%

(a)

Si-OD

Si-OH

0.10%

(b)

(i)

(ii)

(iii)

(iv)

(v)

(vi)

(vii)

(viii)

(i)

(ii)

(iii)

(iv)

(v)

(vi)

(vii)

(viii)

Figure D-6: IRRAS of the effect of D2O at varying temperatures and pressures of D2O

for (a) MFI nanosheets supported on Au(111) and (b) polymorphous bilayer silicate

on Ru(0001). IRRAS is shown for: (i) pre-exposure to D2O at 300 K, UHV, (ii) 10-3

mbar at 473 K for 60 s, (iii) 10-2 mbar at 473 K for 1 h, (iv) 10-1 mbar at 473 K for 1 h,

(v) 1 mbar at 473 K for 1 h, (vi) 3 mbar at 473 K for 1 h, (vii) 3 mbar at 573 K for 1 h,

and (viii) post-exposure to D2O at 300 K, UHV. Spectra in (ii) through (viii) in (a) and

(b) use (i) as a reference. As a result of this, when plotted in % reflectance units,

features pointing up indicate the disappearance of pre-existing species, while features

pointing down show the appearance of new species.

139

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Rati

o o

f S

iOH

:SiO

H3

00

K,

UH

V p

eak

are

as

35003600370038003900

Ref

lect

an

ce

Wavenumber (cm-1)

3736 cm-1

0.05%

300 K, UHV

573 K, UHV

573 K, 3 mbar

300 K, UHV

Pre-H2O

Post-H2O

(a) (b)

Figure D-7: (a) IRRAS of bilayer silicate supported on Ru(0001). Pre-, post-, 573 K

under UHV, and 573 K under 3 mbar H2O for 1 h are specified. (b) Quantification of

the IRRAS peak areas for SiOH, taken relative to the initial amount of SiOH obtained

prior to hydrothermal conditions at 300 K and UHV conditions.

140

A FUNDAMENTAL STUDY ON PHOSPHORUS-CONTAINING ZEOSILS

E.1 Background

In an effort to reduce our greenhouse gas emissions and dependence on fossil

fuels, recent endeavors have focused on upgrading lignocellulosic biomass to fuels

and chemicals.12 Phosphorus-containing zeosils (P-zeosils) have gained much

attention in this pursuit.50,51 In particular, P-BEA and P-SPP were found to convert

biomass-derived dimethylfuran (DMF) to para-xylene with 97% yield. These P-zeosils

were more active and selective than Al-BEA, Zr-BEA, and H3PO4.50 In another report,

P-BEA, P-MFI, and P-SPP were highly selective (>85%) in the conversion of

tetrahydrofuran (THF) to butadiene.51 Although P-zeosils are highly active in biomass

chemistries, we still lack a fundamental understanding in the following areas: (1) the

interaction between phosphorus and the zeosil framework, (2) the strength of the acid

sites in P-zeosils, and (3) the distribution of the P-sites in the presence of steam. In the

present work, we provide insights into each of these areas.

E.2 Interaction Between Phosphorus and the Zeosil Framework

In this section, we investigated the effects of monolayer deposits of P-

molecules, including phosphoric acid (H3PO4), trimethylphosphine (P(CH3)3), and

triethylphosphate ((C2H5)3PO4) on two-dimensional (2-D) covalently saturated

(essentially silanol-free) silicates. These 2-D models are well defined and

Appendix E

141

homogeneous, compared to their 3-D counterparts, and therefore, are an excellent

starting point to determine the interaction between phosphorus and the silica surface.

Prior to phosphorus exposure, the 2-D bilayer silicate was characterized by

IRRAS. The sharp band at 1300 cm-1 in Figure E-1 was indicative of a crystalline

bilayer silicate supported on Ru(0001).44 Note the other regions of the IRRAS (1500 –

4000 cm-1) did not contain any features and are therefore not shown.

Next, we exposed the 2-D bilayer silicate to each phosphorus source. Between

depositions, the bilayer silicate was conditioned by annealing in 1×10-5 mbar O2 at 750

K for 5 min. During deposition, the sample was cooled down between 100 K and 145

K in ultra-high vacuum (UHV) conditions, followed by exposure to the phosphorus

source for a specified amount of time. H3PO4, P(CH3)3, and (C2H5)3PO4 were

deposited at 7.5 × 10-5 Torr for 30 s, 10 s, and 60 s, or 22.5, 7.5, and 45 L,

respectively. H3PO4 and P(CH3)3 were applied for lower amounts of time because of

their higher volatility, and hence greater likelihood to contaminate the IRRAS

chamber, compared to (C2H5)3PO4.

100011001200130014001500

Ref

lect

an

ce

Wavenumber (cm-1)

0.50 %1300 cm-1

Figure E-1: IRRAS spectra for the SiO2 phonons in the bilayer silicate supported on

Ru(0001).

142

We began by studying the interaction between H3PO4 and the bilayer silicate.

H3PO4 is commonly used in the synthesis of phosphorus-containing zeosils.50,51 We

exposed the bilayer silicate to H3PO4 vapor through sublimation of solid H3PO4.

Figure E-2 shows the result of H3PO4 deposition on the bilayer silicate. The modes

corresponding to each wavenumber are provided in Table E-1.

Table E-1: IRRAS features and corresponding modes for H3PO4.

Wavenumber Mode, from Sun et al.167

1673 Water bend

1132 P-OH stretch (sym.)

953 P-OH stretch (anti sym.)

868 O=P=O (anti sym.)

Next, we incrementally increased the temperature of the bilayer silicate to

determine if H3PO4 would remain anchored to the surface. Figure E-3a shows the

deposition of H3PO4 at 141 K. Subsequent spectra were acquired at elevated

temperatures, though the exact values were not recorded. At 250 K, H3PO4 began to

60080010001200140016001800

Ref

lect

an

ce

Wavenumber (cm-1)

1673

1132

953

868

0.50%


H3PO4 at 141 K. The feature at 1300 cm-1 (unlabeled) corresponds to the SiO2

phonons in the bilayer silicate.

143

desorb and had fully desorbed when the temperature was 283 K. Additional

experiments were attempted with H3PO4, but by our third effort, H3PO4 was no longer

depositing, even though the presence of solid H3PO4 was still in our experimental set-

up. The lack of deposition was likely due to the formation of polymeric H3PO4 species

upon water removal during evacuation.168 These polymeric species have a low vapor

pressure and are thereby unable to sublimate and enter the IRRAS chamber.

Given the challenges associated with H3PO4 deposition, we next attempted

dosing P(CH3)3. P(CH3)3 is a reactive and volatile species.169 Figure E-4 shows the

result of P(CH3)3 deposition on the bilayer silicate. The modes corresponding to each

wavenumber are provided in Table E-2.

60012001800240030003600

Ref

lect

an

ce

Wavenumber (cm-1)

141 K

283 K

250 K

2%

1000150020002500300035004000

Ref

lect

an

ce

Wavenumber (cm-1)

270 K

100 K

2%

(b)(a)


(a) H3PO4 at 141 K. (b) H3PO4 fully desorbs at 283 K.

144

Table E-2: IRRAS features and corresponding modes for P(CH3)3.

Wavenumber Mode, from Rumplmayr et al.169

2958

C-H stretching 2897

2814

1435 C-H deformation

1414

We incrementally increased the temperature of the bilayer silicate to determine

if P(CH3)3 would remain anchored to the surface. Figure E-3b shows the deposition of

P(CH3)3 at 100 K. Subsequent spectra were acquired at elevated temperatures. At 270

K, P(CH3)3 had fully desorbed from the surface. The volatile nature of P(CH3)3 made

it difficult to control under UHV conditions. Therefore, we also performed

experiments with the less volatile species, (C2H5)3PO4.

1000150020002500300035004000

Ref

lect

an

ce

Wavenumber (cm-1)

C-H stretching

C-H deformation

1%


P(CH3)3 at 141 K. The feature for the bilayer silicate (1300 cm1) overlaps with the C-

H deformation bands.

145

We selected (C2H5)3PO4 because it has a low vapor pressure and is chemically

similar to H3PO4. Figure E-5 shows the result of (C2H5)3PO4 deposition on the bilayer

silicate. The modes corresponding to each wavenumber are provided in Table E-3. In-

line with our previous attempts, we found that (C2H5)3PO4 fully desorbed at 210 K, as

shown in Figure E-6.

In summary, each of the phosphorus-containing species desorbed from the

bilayer silicate’s surface. The desorption temperatures were within the range of 210 K

and 270 K; however, additional systematic study is needed to verify the exact

desorption temperature for each species. Furthermore, P(CH3)3 and H3PO4 were

difficult to handle under UHV conditions, while (C2H5)3PO4 appeared compatible with

UHV conditions. In the future, if vapor deposition techniques are pursued, we

recommend selecting (C2H5)3PO4, or other phosphate species with a low vapor

pressure for study. In addition to vapor deposition, we recommend wet impregnation

of aqueous phosphoric acid onto the surface, since this techniques is commonly

applied when making phosphorus-containing catalysts.

9001000110012001300140015001600

Ref

lect

an

ce

Wavenumber (cm-1)

991

1052

1108

1169

1259

1369

1483

14451396

1068

10%


(C2H5)3PO4 at 141 K. The feature at 1300 cm-1 (unlabeled) corresponds to the SiO2

phonons in the bilayer silicate.

146

Regarding the surface, in this work, we studied the bilayer silicate; however, it

lacked silanol groups, which are known to serve as anchoring points for metals. Given

this, we recommend performing vapor deposition experiments with surfaces saturated

in silanol groups, such as 2-D MFI nanosheets. MFI nanosheets make up the P-SPP

zeosil-framework, a highly effective catalyst in butadiene and para-xylene

production.50,51

Table E-3: IRRAS features and corresponding modes for (C2H5)3PO4.

Wavenumber (cm-1) Mode, from Kycia et al.170

991 C-C

(P)-O-C

1052 In-phase

1068 Out-of-phase

CH3 rocking

1108 In-plane

1169 Out-of-plane

1259 P=O

CH3 deformation

1369 Symmetric

1445 Asymmetric

CH2

1396 Wagging

1483 CH2 deformation

147

E.3 Strength of the Acid Sites in P-zeosils

In this section, we investigated the strength of the acid sites in phosphorus-

containing zeosils, namely P-BEA and P-SPP, and compared their strength to de-

aluminated BEA, the precursor to P-BEA. The catalysts were prepared as reported by

Cho et al.50 Here, we monitored the adsorption of D2O on the hydroxyl sites at room

temperature by transmission IR. The adsorption profiles, fit to Langmuir isotherms,

showed that the POH sites have a higher propensity to adsorb water than the SiOH

sites.

We began by acquiring transmission IR spectra for P-BEA (Si/P = 27), P-SPP

(Si/P = 27), and de-aluminated BEA (deAlBEA). Although we acquired the full

spectrum (600 – 4000 cm-1) for each zeosil, we emphasize the hydroxyl region (3000 –

4000 cm-1) in Figure E-7. P-BEA and P-SPP contained SiOH and POH sites at 3746

and 3670 cm-1,171 respectively, while de-aluminated BEA contained SiOH groups and

SiOH nests (not labeled, 3600 – 3250 cm-1).

60080010001200140016001800

Ref

lect

an

ce

Wavenumber (cm-1)

110 K

120 K

140 K

160 K

180 K

200 K

210 K

280 K

50%


(C2H5)3PO4 at 110 K. (C2H5)3PO4 fully desorbs at 210 K.

148

Next, we exposed each zeosil to increasing pressures of D2O. Figure E-8 shows

the effect of D2O on de-aluminated BEA. First, de-aluminated BEA was preheated at

623 K to remove any water from its pores. Then, it was cooled to 300 K and exposed

to increasing pressures of D2O. An exchange from SiOH (3732 cm-1) to SiOD (2748

cm-1) occurred. Likewise, SiOH nests became SiOD nests. The top-most spectrum,

labeled (v), was obtained post-D2O exposure. We also performed the same

experiments for P-BEA and P-SPP powders, shown in Figure E-9 and Figure E-10,

respectively. The top-most spectrum corresponds to post-D2O exposure.

Following exposure to D2O, the data (Figure E-8, Figure E-9, and Figure E-10)

were fit with a Langmuir-type equation (Eq. E-1).

30003250350037504000

Ab

sorb

an

ce (

a.u

)

Wavenumber (cm-1)

P-BEA

P-SPP

deAlBEA

(Si/P = 27)

(Si/P = 27)

3670 cm-1

3746 cm-1

20%

Figure E-7: P-BEA (Si/P = 27), P-SPP (Si/P = 27), and de-aluminated BEA

(deAlBEA) spectra, acquired at 300 K. Catalysts were pre-heated at 623 K under

vacuum (1×10-5 Torr) for 12 h. Spectra were acquired upon cooling to 300 K.

149

N = NML

KC

KC+1 Eq. E-1

The fitting parameters are NML and K. NML represents the total number of sites capable

of reacting with D2O (a fraction of all -OH groups), and K is the equilibrium constant.

The decrease in band height is proportional to the amount of D2O adsorbed on the

SiOH and POH sites. This decrease in height, N, is plotted as a function of increasing

D2O pressure (Torr), as shown in Figure E-11. C is the pressure of D2O (Torr). Table

E-4 provides the fitting parameters (NML and K). We found that the POH sites have a

higher equilibrium constant (K) compared to SiOH sites, indicating these sites have a

higher propensity to adsorb water. This coincides with the hypothesis that the POH

sites are the active in biomass-upgrading reactions.50

65013001950260032503900

Ab

sorb

an

ce (

a.u

.)

Wavenumber (cm-1)

3732 cm-1

3472 cm-1

2748 cm-1

2604 cm-1

(i)

(ii)

(iii)

(iv)

(v)

20%

Figure E-8: De-aluminated BEA was pre-heated at 623 K under vacuum (1×10-5 mbar)

for 12 h, followed by exposure to varying levels of D2O at 300 K for 1 h. The

corresponding pressures include: (i) 3.8×10-7 Torr, pre-D2O, (ii) 4.0×10-4 Torr D2O,

(iii) 2.3×10-1 Torr D2O, (iv) 2.3 Torr D2O, and (v) 3.9×10-4 Torr, post-D2O. SiOH

(3732 cm-1) gradually exchanges to SiOD (2748 cm-1). Likewise, SiOH nests (3472

cm-1) gradually exchange to SiOD nests (2604 cm-1).

150

600110016002100260031003600

Ab

sorb

an

ce (

a.u

.)

Wavenumber (cm-1)

3745 cm-1

3670 cm-1

2760 cm-1

2706 cm-1

20%

Figure E-10: P-SPP was pre-heated at 623 K under vacuum (1×10-5 mbar) for 12 h,

followed by exposure to varying levels of D2O at 300 K for 1 h. The corresponding

pressures include: (i) 3.4×10-7 Torr, pre-D2O, (ii) 3.8×10-3 Torr D2O, (iii) 3.2×10-2

Torr D2O, (iv), 1.4×10-1 Torr D2O (v) 1.0 Torr D2O, (vi) 2.0 Torr D2O, (vii) 3.0 Torr

D2O, (viii) 4.4 Torr D2O, (ix) 5.0 Torr D2O, (x) 6.3 Torr D2O, (xi) 4.0×10-4 Torr, post-

D2O . SiOH (3745 cm-1) gradually exchanges to SiOD (2760 cm-1). POH (3670 cm-1)

gradually exchanges to POD (2706 cm-1).

600110016002100260031003600

Ab

sorb

an

ce (

a.u

.)

Wavenumber (cm-1)

3745 cm-1

3666 cm-1

2756 cm-1

2692 cm-1

(i)

(ii)

(iii)

(iv)

(v)

(vi)

(vii)

20%

Figure E-9: P-BEA was pre-heated at 623 K under vacuum (1×10-5 mbar) for 12 h,

followed by exposure to varying levels of D2O at 300 K for 1 h. The corresponding

pressures include: (i) 2.4×10-7 Torr, pre-D2O, (ii) 11.3×10-3 Torr D2O, (iii) 1.9×10-1

Torr D2O, (iv) 4.0 Torr D2O, and (v) 5.6 Torr D2O, (vi) 6.7 Torr D2O, (vii) 7.5×10-4

Torr, post-D2O. SiOH (3745 cm-1) gradually exchanges to SiOD (2756 cm-1). POH

(3666 cm-1) gradually exchanges to POD (2692 cm-1).

151

Table E-4: The fitting parameters (NML and K) and R2 value for the hydroxyl sites in

de-aluminated BEA, P-BEA, and P-SPP.

Hydroxyl (-OH) Site NML

K R2

De-Al BEA

Si-OH 0.28 10 0.933

P-BEA Si-OH

0.25 0.22 0.944

P-SPP Si-OH

0.65 0.10 0.967

P-SPP P-OH

0.018 5 1

P-BEA P-OH

0.023 0.29 0.994

In summary, we exposed de-aluminated BEA, P-BEA, and P-SPP to D2O. We

found that the POH sites in P-BEA and P-SPP had a higher propensity to adsorb D2O

than the SiOH sites. This implies the active site of P-zeosils likely contains the POH

sites, as they are more reactive than SiOH sites. In the future, we recommend

0

0.05

0.1

0.15

0.2

0.25

0.3

0 2 4 6 8 10

Ba

nd

Hei

gh

t D

ecr

ease

(N

)

D2O Pressure (Torr)

De-Al BEA SiOH

P-BEA SiOH

P-SPP SiOH

P-SPP POH

P-BEA POH

300 K

Figure E-11: The band height decrease (N, y-axis), obtained from the transmission IR

spectra for the SiOH and POH groups in P-BEA, P-SPP, and de-aluminated BEA,

were plotted as a function of increasing D2O pressure (Torr). The data were fit with a

Langmuir-type equation.

152

performing similar reaction studies with the reactants, DMF and THF, and studying

the interaction between these reactants and the framework at reaction conditions.

E.4 Distribution of P-Sites During Steaming

In this section, we monitored the effects of steam on P-BEA (Si/P = 3) by

transmission IR. The purpose of these experiments was to determine if changes in

speciation occur under reaction conditions. Here, the catalysts contain a higher

fraction of phosphorus, compared to the previous section. On-going work in the field

is aiming to understand the effects of a high phosphorus content on reaction

chemistries.

Prior to obtaining measurements under steam, P-BEA underwent pre-treatment

at 673 K for 4 h under vacuum. A spectrum for P-BEA was acquired prior to H2O

exposure (Figure E-12).

35003600370038003900

Ab

sorb

an

ce (

a.u

.)

Wavenumber (cm-1)

0.5%

65013001950260032503900

Ab

sorb

an

ce (

a.u

.)

Wavenumber (cm-1)

2%

Post-exposure to H2O, 523 K

Pre-exposure to H2O, 523 K

10 Torr H2O, 523 K, 1 h

(b)(a)

POH

3665 cm-1SiOH

3740 cm-1


during pre-exposure to H2O at 1×10-5 mbar (pink), 10 Torr H2O (blue), and post-

exposure to H2O at 1×10-5 mbar (orange). All spectra were acquired at 523 K. (b)

Zooming in on the hydroxyl region of the IR spectra for P-BEA, SiOH and POH

groups appear at 3740 cm-1 and 3665 cm-1, respectively. Prior to acquisition, the

catalyst sample was heated under vacuum (1×10-5 mbar) at 673 K for 4 h.

153

Then, P-BEA was exposed to 10 Torr H2O at 523 K for 1 h. These conditions

are comparable to reaction conditions. We also acquired data as a function of time,

shown in Figure E-13. Unfortunately, we were unable to draw any major conclusions

from these experiments, as the spectra did not show any differences pre-, post-, and

during steaming.

65013001950260032503900

Ab

sorb

an

ce (

a.u

.)

Wavenumber (cm-1)

1 hr

15 min

30 min

45 min

2%

35003600370038003900

Ab

sorb

an

ce (

a.u

.)

Wavenumber (cm-1)

POH

3665 cm-1

SiOH

3740 cm-1

0.5%

(b)(a)


during pre-exposure at 523 K and 10 Torr H2O as a function of time. (b) Zooming in

on the hydroxyl region of P-BEA, SiOH and POH groups appear at 3740 cm-1 and

3665 cm-1, respectively. Prior to acquisition, the catalyst sample was heated under

vacuum (1×10-5 mbar) at 673 K for 4 h.

154

PERMISSIONS FOR REPRINT

Parts of Chapter 1 and all of Chapter 2 were reproduced with permission from

(https://doi.org/10.1021/acscatal.9b01853) Copyright © 2019 American Chemical

Society.

Chapter 3 was reproduced from work by Norton et al.89 with permission from the

Royal Society of Chemistry Advances (https://doi.org/10.1039/C8RA03088J).

Chapter 4 was reprinted with permission from

(https://doi.org/10.1002/cssc.201901838) Copyright © 2019 John Wiley and Sons.

Chapter 5 was reproduced with permission from the Journal of Physical Chemistry C,

in press. Unpublished work copyright 2020 American Chemical Society.

Appendix F

https://doi.org/10.1021/acscatal.9b01853

https://doi.org/10.1039/C8RA03088J

https://doi.org/10.1002/cssc.201901838

155

Figure F-1: Permission for Chapter 1 and Chapter 2

156

157

Figure F-2: Permission for Chapter 4

Documents

homogeneous and heterogeneous catalysis for biomass