Optimized Cellulose Nanocrystal Organocatalysts Outperform Silica-Supported Analogues: Cooperativity, Selectivity, and Bifunctionality in Acid-Base Aldol Condensation Reactions

Article abstract of DOI:10.1021/acscatal.8b05180

Cellulose nanocrystals (CNCs) are demonstrated as effective, ordered supports for cooperative acid-base heterogeneous organocatalysis, offering an alternative to typical silica supports. CNC catalyst surface chemistry is optimized through quantitative control of the loadings of carboxylic acids, primary amines, and sulfate half-esters, as characterized by elemental analysis, conductometric titration, and FT-IR spectroscopy. Catalysts are evaluated in the liquid phase aldol condensation of 4-nitrobenzaldehyde or furfural with acetone. Carboxylic acids are effective cooperative acid partners in CNC organocatalysts, and site-specific activity is strongly correlated with the COOH:NH₂ ratio. Partial sulfate half-ester removal, high acid/base ratios, and use of unprotected diamines in the catalyst synthesis lead to optimized CNC catalyst function (site-time yield = 1.0 × 10^-4 s^-1). High selectivities to dehydrated aldol products (>80%) are achieved due to the acid content of the CNC catalysts. CNC catalysts outperform analogous SBA-15-supported aminosilica catalysts in regard to both activity and selectivity. Crystalline surface structures and ordered chemical functionalization in CNCs appear advantageous for precise design and control of bifunctional acid-base cooperative catalysts.

Full text of DOI:10.1021/acscatal.8b05180

Subscriber access provided by ECU Libraries
Article
Optimized cellulose nanocrystal organocatalysts outperform
silica-supported analogues: cooperativity, selectivity, and
bifunctionality in acid–base aldol condensation reactions
Nathan C Ellebracht, and Christopher W Jones
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b05180 • Publication Date (Web): 01 Mar 2019
Downloaded from http://pubs.acs.org on March 4, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 43
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
2
3
4
5
Optimized cellulose nanocrystal
organocatalysts outperform silica-supported
analogues: cooperativity, selectivity, and
bifunctionality in acid–base aldol condensation
reactions
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
7
8
9
Nathan C. Ellebracht, Christopher W. Jones*
School of Chemical & Biomolecular Engineering, Georgia Institute of Technology, 311
Ferst Dr., Atlanta, Georgia 30332-0100, United States
*cjones@chbe.gatech.edu
1
1
0
1
Keywords
12
Cellulose nanocrystals, heterogeneous organocatalysis, aldol condensation, furfural, acid-
base, cooperative catalysis, site pairing, optimization
1
1
1
1
1
1
3
4
5
6
7
8
Abstract
Cellulose nanocrystals (CNCs) are demonstrated as effective, ordered supports for
cooperative acid-base heterogeneous organocatalysis, offering an alternative to typical silica
supports. CNC catalyst surface chemistry is optimized through quantitative control of the
loadings of carboxylic acids, primary amines, and sulfate half-esters, as characterized by
19
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
0
1
elemental analysis, conductometric titration, and FT-IR spectroscopy. Catalysts are evaluated
in the liquid phase aldol condensation of 4-nitrobenzaldehyde or furfural with acetone.
Carboxylic acids are effective cooperative acid partners in CNC organocatalysts, and site-
22
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
3
4
specific activity is strongly correlated with the COOH:NH ratio. Partial sulfate half-ester
2
removal, high acid:base ratios, and use of unprotected diamines in the catalyst synthesis lead
-4
-1
25
to optimized CNC catalyst function (site-time yield = 1.0 × 10 s ). High selectivities to
dehydrated aldol products (>80 %) are achieved due to the acid content of the CNC catalysts.
CNC catalysts outperform analogous SBA-15-supported aminosilica catalysts in regard to
both activity and selectivity. Crystalline surface structures and ordered chemical
functionalization in CNCs appear advantageous for precise design and control of bifunctional
acid-base cooperative catalysts.
2
2
6
7
28
2
3
9
0
ACS Paragon Plus Environment

Page 3 of 43
ACS Catalysis
1
2
3
4
3
1
2
1 Introduction
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
Organocatalysis in organic chemistry and pharmaceutical synthesis has been
1
33
substantially developed in recent decades. Organocatalytic transformations toward fine
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chemicals have often involved enantioselective amine catalysts like proline,2–4 but the
3
3
4
5
5
breadth of organocatalyst-mediated reactions is rapidly expanding. Heterogeneous
36
organocatalysts allow for immobilization of multiple catalytic species in a single material for
cascade and/or cooperative catalysis.6–8 Silane-modified mesoporous silicas are a major class
of heterogeneous organocatalysts studied to date and have been developed with a variety of
functionalities, commonly utilizing organic acids and bases.9–11 These acid and/or base
catalysts have been used for various C–C bond forming reactions including aldol, nitroaldol,
and Knoevenagel condensations as well as Michael addition and deacetalization reactions.
The aldol condensation (AC) of an aldehyde with acetone is a common acid-base
3
3
3
4
4
4
7
8
9
0
1
2
43
cooperatively catalyzed test reaction with an established mechanism,12–14 and many aspects
of the active site have been investigated. Acid sites, including sulfonic,6,12,15,16
4
4
4
5
phosphoric,12,17 and carboxylic acids,12,17,18 as well as organic bases like chiral and various
alkyl13,20 amines have been incorporated in silica-supported organocatalysts. Weakly
Brønsted acidic silanols have been established as effective cooperative acid partners in
aminosilica catalysts due to their abundance and insignificant base-poisoning effects.8,18,20
However, weak acids of greater strength may be more effective in some cases by reducing
reaction activation energy, as was shown for butanal self-condensations.21
19
46
4
4
4
5
5
5
7
8
9
0
1
2
The mild reaction conditions and acid-base chemistries typical of supported
2
2
organocatalysts are appropriate for certain key transformations of biomass. Upgrading of
biomass-derived furfural via aldol condensation with ketones and subsequent
53
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
5
4
5
hydrodeoxygenation of condensation products has been studied with a number of inorganic
heterogeneous catalysts including zeolites, layered hydroxides, mixed oxides, and noble
metals.23–26 Despite their relevance, heterogeneous organocatalysts have been sparsely
56
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
27
5
5
7
8
utilized for transformations of biomass. Acid-catalyzed conversion of sugars to furans and
1
6
condensations of sugar-derived furans were shown with supported sulfonic acid catalysts.
Nelson et al utilized a bifunctional Pd and secondary amine mesoporous silica nanoparticle
catalyst for a cascade transformation of furfuryl alcohol to furfural and the aldol condensation
product, however with low yields of the final product.28 Supported organocatalysts are
promising yet understudied as components of multifunctional catalysts for cascade furfural
upgrading reactions.
59
6
6
6
6
6
6
0
1
2
3
4
5
Spacing and pairing of acid and base species are key in cooperative catalysis. Polar
protic grafting solvents, protecting groups, or multi-step functionalizations may aid in
dispersion of grafted aminosilanes.29–32 Moderate loadings with neither site-isolated nor
66
32
6
6
7
8
overcrowded aminosilanes was determined to be optimal for the Henry reaction. Without
appropriate grafting methods, aminosilane clustering may occur even at very low loadings.33
Dispersion and acid-base site-pairing in multifunctional silica-supported organocatalysts is
even more difficult to control.15,34 Some control over reaction environment has been
demonstrated via silica imprinting with protected aminosilanes.35,36 Random grafting and the
amorphous nature of the oxide supports typical of this material class are limiting factors in
the design of cooperative active sites.
69
7
7
7
7
7
7
0
1
2
3
4
5
Cellulose nanocrystals (CNCs) are an emergent class of advanced bioproducts
3
7
derived from various cellulose sources with the potential to be produced at scale. CNCs are
76
high aspect ratio, high specific surface area rods (100–300 nm in length) of crystalline
ACS Paragon Plus Environment

Page 5 of 43
ACS Catalysis
1
2
3
4
7
7
8
cellulose with typically rectangular cross sections (5–10 nm widths).38 The ordered,
hydroxyl-rich surfaces of CNCs are readily functionalizable, modification of which has been
widely studied and reviewed.39,40 TEMPO-mediated oxidation, perhaps the most common
chemical modification of CNCs, selectively oxidizes the primary (C6) surface hydroxyls to
carboxylic acids, which are regularly ordered due to the crystalline nature of cellulose.41,42
CNCs have been used as catalyst supports primarily through reduction of noble metal
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
7
79
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
8
0
1
82
43
8
8
8
8
8
8
3
4
5
6
7
8
salts into nanoparticles on oxidized surfaces and are studied with test reactions under mild
conditions. The slight chiral twist of CNC rods has been shown to promote enantiomeric
4
4
45
selectivity in Pd-catalyzed hydrogenation and amino acid hydrolysis. Several studies have
utilized CNCs as direct organocatalysts. Hydrolysis reactions have been performed with
unmodified CNCs45,46 as well as TEMPO-oxidized nanofibers. Kanomata et al performed
aldol condensations catalyzed by a combined heterogeneous-homogeneous system of
47
4
8
89
oxidized nanofibers and homogeneous proline. More recently, our previous work was an
initial demonstration of acid-base bifunctional CNCs as cooperative heterogeneous
organocatalysts for AC test reactions.49
9
9
0
1
92
There is a large body of work refining and probing aspects of aminosilica
organocatalyst materials and surface chemistry for C–C bond forming reactions. In contrast,
9
9
9
9
9
9
3
4
5
6
7
8
4
9
only our prior work explores similar issues on nanocellulose. The present study greatly
expands upon the preliminary acid-base CNC organocatalyst design established in our
previous work to explore the surface chemistry and structure of CNCs as they pertain to this
catalysis. Cellulose nanocrystal-supported catalysts are synthesized by TEMPO-mediated
oxidation and subsequent amine functionalization by amide coupling of diamines to a portion
of the surface carboxylic acids. Sulfate half-ester, carboxylic acid, and amine surface species
99
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
00
01
are quantitatively varied and characterized in an array of CNC catalysts. Coupling of
unprotected diamines as well as mono-boc protected diamines is used to probe the effects of
diamide formation on active site availability. Catalysts are tested in batch aldol condensation
reactions of acetone with a test substrate (4-nitrobenzaldehyde) as well as furfural and
compared via activity and selectivity to dehydrated products. Analogous mesoporous SBA-
15-supported organocatalysts are synthesized and directly compared the CNC-supported
catalysts developed here.
102
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
03
04
105
1
1
1
1
1
1
06
07
08
09
10
11
2 Materials and Methods
2.1 Cellulose nanocrystal catalyst preparation
Cellulose nanocrystal (CNC) acid-base organocatalysts were synthesized by chemical
modification of CNCs produced by sulfuric acid hydrolysis and purchased from the
University of Maine in concentrated aqueous suspension (~10 wt%). CNCs were generally
purified by dialysis against deionized water and stored in aqueous suspension at 4 °C.
Cellulose nanocrystals with as-received (~0.9 wt% S), partially removed, or significantly
removed sulfate half-ester (SE) content were used as starting materials for catalyst
preparation. For partial SE removal, an aqueous suspension of CNC (0.5 wt%) and HCl (50
mM) was magnetically stirred at 80 °C for 24 h. For significant SE removal, a similar
suspension with increased HCl (250 mM) was magnetically stirred at 100 °C for 24 h. The
cooled suspensions were partially neutralized to pH ~4 with NaOH before purification by
dialysis and concentration by ultrafiltration.
112
1
1
13
14
115
1
1
1
1
1
1
16
17
18
19
20
21
4
1
These CNCs were subsequently oxidized by an oft-reported procedure mediated by
122
(2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO) with which low or high extents of
ACS Paragon Plus Environment

Page 7 of 43
ACS Catalysis
1
2
3
4
1
1
23
24
oxidation were achieved by varying the NaClO content. To CNC suspensions (2 wt% in
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
H O), TEMPO (0.1 g/g
dry
CNC), NaClO (less oxidation: 0.42; more oxidation: 1.9 g/gdry
2
125
CNC), and NaBr (1 g/gdry CNC) were added and dissolved with stirring. The mixture was
stirred at room temperature for 4 h, during which the pH was continually adjusted to between
10 and 11 with NaOH (1 N) until stabilization. NaCl was then added to flocculate the
oxidized CNCs, and the product was washed by centrifugation and decanting (three times
with 0.5 M NaCl; two times with 0.1 M HCl; once with DI water) before final purification
by dialysis. Oxidized CNCs were further modified by aqueous amide coupling with 1,3-
diaminopropane (DAP, >99%, Sigma-Aldrich) or n-boc-1,3-diaminopropane (bocDAP). The
singly protected diamine (bocDAP) was synthesized by previously reported methods (Figure
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
26
27
128
1
1
1
1
1
1
29
30
31
32
33
34
5
0
S1). DAP or bocDAP (five molar equivalents with respect to CNC carboxylic acid content),
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (1.5 eq, >99%, Sigma-
Aldrich), and n-hydroxysuccinimide (1.5 eq, 98%, Sigma-Aldrich) were added to
135
1
1
36
37
suspensions of oxidized CNCs (1–2 wt% in H O). The pH was adjusted with HCl to 7.5–8,
2
and the mixture was stirred overnight at room temperature prior to purification. CNCs
functionalized with bocDAP were hydrothermally deprotected by stirring aqueous
suspensions (0.5 wt%) at 100 °C for 7 h.
138
1
1
1
1
1
1
39
40
41
42
43
44
2.2 SBA-15 catalyst preparation
SBA-15 aminosilica catalysts were prepared according to previously reported
methods.18 To synthesize mesoporous SBA-15, pluronic P123 copolymer (24.0 g, Sigma-
Aldrich) was dissolved in deionized water (636 mL) and hydrochloric acid (120. mL, 12.1
M). Tetraethylorthosilicate (TEOS, 46.26 g, 98%, Sigma-Aldrich) was added and the
solution was stirred at 40 °C for 20 h. The stir bar was removed, and the temperature was
145
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
46
47
increased to and held at 100 °C for 24 h. The solids were filtered and washed with deionized
water (4 L) followed by drying overnight at 75 °C. Finally, the material was calcined (RT to
-1
-1
148
200 °C at 1.2 °C min , 200 °C for 1 h, 200 to 500 °C at 1.2 °C min , 550 °C for 12 h, cool
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
49
50
to RT) and dried overnight (110 °C, 10 mTorr) before storage or use.
Acid-base aminosilica catalysts were prepared by silane grafting. Dried SBA-15 (1.5
151
g) was stirred under inert atmosphere (N ) in toluene (37.5 mL) with the desired silane (199
2
1
1
1
1
1
1
52
53
54
55
56
57
mg, 3-aminopropyltriethoxysilane, APS, Gelest) for 24 h. Subsequently, deionized water (30
µL) was added and the mixture was stirred at 80 °C for 24 h followed by filtration and
washing with toluene, hexanes, and ethanol. The washed powder was dried overnight (110
°C, 10 mTorr) before storage or use. An aminosilica catalyst incorporating both grafted
carboxylic acid and primary amine silanes was also synthesized. A protected acid silane, t-
butyl 2-methyl-3-(triethoxysilyl)-propanoate (191 mg, TBSP, synthesized as previously
1
8
158
reported, Figure S2), was similarly grafted to bare SBA-15 (1.5 g). APS (110 mg) was
subsequently grafted to this intermediate material (1.2 g) with the above approach. The tert-
butyl protecting group was removed by heating to 270 °C for 12 h under flowing N2
exhausted through a bubbler. The APS-only catalyst is denoted by SBA-APS, and the dual
grafted catalyst is referred to as SBA-COOH-APS.
1
1
59
60
161
62
1
ACS Paragon Plus Environment

Page 9 of 43
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
63
64
Figure 1. Synthesis scheme for cellulose nanocrystal- and SBA-15-supported acid-base
organocatalysts. Partial diamide crosslinking and incomplete deprotection are shown for
DAP and bocDAP CNC-based catalysts, respectively; the relative quantities of the surface
species shown are not quantitatively accurate. Italicized steps were performed only for a
subset of the materials synthesized.
1
65
166
1
1
67
68
1
69
2.3 Catalyst characterization
170
Freeze-dried CNC materials and silica materials were analyzed by combustion
elemental analysis (EA) performed by Atlantic Microlab, Inc. for elemental content (some
or all of: C H N S). Conductometric titration was performed to quantify the carboxylic acid
content of oxidized CNCs. Simultaneous pH and conductivity measurements were recorded
with a Hanna Instruments benchtop meter. Suspensions of CNCs (~0.1 wt%) were acidified
1
1
71
72
173
1
1
1
74
75
76
(pH ~3), purged with N (30 min), and titrated with NaOH (0.100 N) to pH >10. Acid content
2
was determined from the low slope portion of the conductivity curve around pH 4–8. Freeze-
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
77
78
dried CNC and silica samples were pelletized with KBr and characterized by Fourier
transform-infrared (FT-IR) transmission spectroscopy with a Nicolet iS10 Spectrometer.
179
BET surface area and pore volume of silica materials were determined from N physisorption
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
80
81
isotherms collected with a Micromeritics TriStar II instrument. Thermogravimetric analysis
(TGA) was performed with a Netzsch STA409 instrument to determine grafted silane
182
loadings in SBA-15 materials. Powder samples were analyzed during heating from 30 to 900
-1
1
1
83
84
°C at a ramp rate of 10 °C min under combined air (90 sccm) and N (30 sccm) flow.
2
Representative I훽 cellulose crystal structures were generated with the Cellulose-builder
5
1
185
toolkit and modified in Jmol.
1
1
1
86
87
88
2.4 Aldol condensation reactions and product analysis
CNC and SBA-15 catalysts were tested in batch aldol condensation reactions of 4-
nitrobenzaldehyde or furfural with acetone repeated at least three times. Reactions were
performed in condenser-fitted two-neck 25 mL round bottom flasks with 10 mol % catalyst
(calculated for CNC catalysts: active site loading = total diamine content = ½ nitrogen
content; for SBA-15 catalysts: active site loading = total amine content = nitrogen content)
at 60 °C in 50:50 v%:v% acetone:acetonitrile (4 mL) suspensions containing either 4-
nitrobenzaldehyde or furfural as the starting material and 1,4-dimethoxybenzene as an
internal standard (all 50 mM). Reactions with furfural were performed under inert (N2)
189
1
1
90
91
192
1
1
93
94
4
9
195
conditions. CNC catalyst processing was described in a previous work. Briefly, a volume
(typically 2–5 mL) of a stored aqueous suspension containing the desired mass of CNC was
dialyzed against 50:50 v%:v% acetone:acetonitrile (1 L) overnight. The resulting gel was
carefully transferred and washed with this solvent mixture three times before being
transferred to a reaction flask in a 3 mL suspension. For silica catalysts, the material was
1
1
1
96
97
98
199
ACS Paragon Plus Environment

Page 11 of 43
ACS Catalysis
1
2
3
4
2
2
00
01
dried overnight (100 °C, 10 mTorr) and combined with 3 mL of the solvent mixture in the
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
flask. The reaction vessel was sealed, purged with N (for reactions of furfural), and heated
2
202
to 60 °C with stirring before a solution (1 mL) of the starting material and internal standard
(200 mM in acetone:acetonitrile) was injected. Aliquots (50–100 µL) were taken via needle
at regular intervals and analyzed by flame ionization detector with a Shimadzu gas
chromatograph. Reactions were cooled after 6 h and final mixtures were collected by
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
03
04
205
2
2
2
2
2
2
06
07
08
09
10
11
filtration. Collected reaction mixtures were concentrated by rotary evaporation and dissolved
1
in DMSO-d for analysis via H NMR with a Bruker Avance III 400 MHz NMR.
6
3 Results and Discussion
3.1 CNC surface chemistry
The surface functionalization methods used result in CNC surfaces with an array of
surface species including: 1) cellulose hydroxyl groups, 2) sulfate half-esters (SEs), 3)
carboxylic acids (in the C6 position of anhydroglucopyranose cellulose monomers), and 4)
amide-coupled diamines with free primary amines. An array of cellulose nanocrystal-
supported acid-base organocatalysts was synthesized by controlling the quantities of these
surface species as well as the type of diamine used for amine functionalization (Figure 1).
The effects of each on resulting catalytic function were probed by varying aspects of the
surface chemistry. The initial quantity of surface hydroxyls is defined by cellulose crystal
212
2
2
13
14
215
2
2
2
2
2
2
16
17
18
19
20
21
5
2
structure and geometry. Sulfate half-ester and coupled diamine contents of a given material
were calculated from S and N wt % from elemental analysis. The carboxylic acid contents of
oxidized CNCs were determined prior to amide coupling by conductometric titration.
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
22
23
All surface species originate from the exterior hydroxyl groups of cellulose
nanocrystals, initially ~4.3 mmol OH per gCNC for Forest Products Laboratory CNCs with
cross-sectional widths of ~5 nm.40,49,52 Unmodified CNC surfaces are slightly decorated with
sulfate half-esters (0.28 mmol gCNC-1) resulting from production by sulfuric acid hydrolysis.
Partial and substantial hydrolytic removal of SEs with HCl resulted in 36–41 and 89 %
reductions in CNC SE content, respectively, consistent with similar procedures from the
literature.53,54 SEs removed by acid hydrolysis result in an additional free surface hydroxyl
but do not substantially change the overall hydroxyl surface content. Substantial CNC
flocculation occurred when SEs were nearly entirely removed (CNC-DS), noted by increased
suspension viscosity and opacity.
224
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
25
26
227
2
2
2
2
2
2
28
29
30
31
32
33
CNC starting materials of various SE contents were oxidized under low or high
NaClO conditions, resulting in low (0.9–1.0 mmol COOH gCNC-1) or high (1.3) initial
234
carboxylic acid content (Table 1). TEMPO-mediated oxidation selectively oxidizes C6
surface hydroxyl groups, which are regularly spaced due to sample crystallinity.42,52 The
2
2
35
36
initial surface coverage of C6 hydroxyls is one third of the total hydroxyl content, about 1.4
mmol gCNC-1, suggesting approximately complete oxidation of these hydroxyls for “high”
oxidation samples. These carboxylic acids serve as the grafting sites for the amine
functionality achieved through aqueous EDC amide coupling with diamines which followed
oxidation.
237
2
2
2
2
2
2
38
39
40
41
42
43
Amide coupling was performed with either of two diamines: 1,3-diaminopropane
(DAP) and n-boc-1,3-diaminopropane (bocDAP). A singly-boc-protected diamine was
utilized to prevent two-site amide coupling of single diamine molecules, which could reduce
free amines and result in CNC cross-linking, an approach that has been employed
244
ACS Paragon Plus Environment

Page 13 of 43
ACS Catalysis
1
2
3
4
5
5
2
2
45
46
previously. Our previous work demonstrated effective catalysis with catalysts synthesized
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
similarly with 1,3-diaminopropane and gave evidence of a relatively low degrees of cross-
4
9
247
linking. Diamine loadings resulting from amide coupling are shown in Table 1. CNC
catalysts are denoted by the degree of desulfation of the starting material (none: s; partial: ds;
high: DS), the extent of oxidation (low: cooh; high: COOH), and the diamine used for
amidation (DAP; deprotected bocDAP: dbDAP). A DAP-functionalized CNC sample was
also subjected to the hydrothermal treatment used for deprotection (Section 2.1) and is
denoted by -HT. Amide coupling typically resulted in diamine loadings of around 0.8 mmol
gCNC-1 regardless of the quantity of available carboxylic acids or diamine type (DAP or
bocDAP) with the exception of the highly desulfated material, for which diamine coupling
resulted in 0.52 mmol bocDAP gCNC-1. This limit suggests possible steric or electrostatic
prevention of further amidation and a practical limit for functionalization density with these
diamines.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
48
49
250
2
2
2
2
2
2
51
52
53
54
55
56
257
2
2
58
59
Removal of the boc protecting group was achieved with simple hydrothermal
processing. Numerous methods for boc deprotection were screened, including TFA, HCl,
5
6
260
NaI, trimethylsilyl iodide-catalyzed approaches. Deprotection was primarily tracked by FT-
IR spectroscopy. Effective cleavage of the boc carbamate group consistently cleaved some
of the amide bonds tethering the diamines to the CNC surfaces as well, leading to a tradeoff
in deprotection and loss of total amine functionality. The 7 h hydrothermal deprotection
approach ultimately utilized removed 30–40 % of the total diamines which liberated
additional carboxylic acid groups. The total amine content of the hydrothermally treated
DAP-functionalized CNC was also reduced by 30 % (Table 1).
2
2
2
2
2
2
61
62
63
64
65
66
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
67
68
Table 1. Quantification of relevant surface species and COOH:NH ratio for the array of
2
CNC catalysts synthesized.
Sulfate half-
Initial
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
b
c
d
a
Initial COOH Final COOH
d
Final diamine COOH:NH
2
Sample
esters (mmol
-1
-1
diamine
-1
e
-1
(mmol gCNC ) (mmol gCNC
)
-1 (mmol gCNC
)
ratio
gCNC
)
(mmol gCNC
)
CNC-s (unmodified)
CNC-ds
0.28
0.17
0.03
0.22
0.14
0.03
0.12
0.24
0.13
0.12
-
-
-
-
-
-
-
-
-
-
CNC-DS
-
-
-
-
-
CNC-s-cooh-dbDAP
CNC-ds-cooh-dbDAP
CNC-DS-cooh-dbDAP
CNC-ds-COOH-dbDAP
CNC-s-cooh-DAP
CNC-ds-COOH-DAP
0.92
0.93
1.0
1.3
0.95
1.3
1.3
0.38
0.42
0.57
0.79
[0.14]
[0.50]
[0.74]
0.76
0.85
0.52
0.90
-
0.54
0.51
0.43
0.51
0.80
0.80
0.56
0.70
0.82
1.3
1.5
0.18
0.60
1.3
-
CNC-ds-COOH-DAP-HT
0.80
a
2
2
271
2
2
2
2
2
2
69
70
Sulfate half-ester content is assumed to be equivalent to elemental S content
b
Determined by conductometric titration prior to amide coupling
c
Residual COOH determined by assuming a 1:1 conversion of COOH to amide surface
attachments with total diamine content. This is a maximum residual COOH value for DAP-
72
73
74
75
76
77
functionalized materials [in brackets] as some crosslinking would convert two COOH per diamine
and further reduce residual acid content.
d
Total diamine content determined by elemental N content (2 N : 1 diamine). The initial content
refers to the loading prior to deprotection or hydrothermal (HT) treatment. If no HT treatment after
to amidation was performed, no initial loading is listed.
e
278
This ratio assumes one free primary amine (NH ) per surface-coupled diamine. Incomplete
2
2
2
2
2
79
80
81
82
deprotection (dbDAP) would reduce the effective free primary amine content and partial
crosslinking (DAP) would decrease both residual COOH and free primary amine contents.
Analysis of FT-IR spectra of oxidized and DAP-functionalized CNCs was described
283
in our previous work.49 Full FT-IR spectra are shown in Figure S4. Chemical species
resulting from oxidation and amidation are particularly evident in the 1500–1800 cm-1
2
2
84
85
spectral range. Numerous IR signals, many overlapping, occur in this range: the carboxylic
-1
-
-1
286
acid C=O stretch (1735–1740 cm ), the deprotonated COO C=O stretch (1605 cm ), the
ACS Paragon Plus Environment

Page 15 of 43
ACS Catalysis
1
2
3
4
-
1
-1
2
2
87
88
amide I C=O stretch (1655–1665 cm ), N–H bending of primary amines (~1600 cm ), and
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
-1
secondary amide II peaks (1535–1540 cm ). Importantly, the carbamate C=O of a boc-
-1
289
protected amine (1685–1695 cm ) was used to qualitatively track the extent of deprotection
(Figure 2). With deprotection, this peak as well as the amide II stretch were significantly
reduced. However, the 7 h hydrothermal deprotection used for all boc-protected samples
resulted in some residual carbamate C=O intensity, qualitatively indicative of incomplete
deprotection.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
90
91
292
2
2
2
2
2
2
93
94
95
96
97
98
Longer times or increased temperature could more fully remove boc protecting
groups but would result in significant losses of total amine content. A previous work utilizing
similar functionalization chemistries did not discuss or confirm tethering amide cleavage
5
5
after deprotection with TFA. FTIR spectra of deprotected bocDAP catalysts otherwise
closely resemble those of the DAP-functionalized catalysts. Additionally, a reduction in
relative intensity of the minor amide II peak after hydrothermal treatment of the DAP catalyst
suggests fewer amide bonds are present, possibly attributable to a reduction in crosslinked
diamide species. This would indicate that the free primary amine loading is closer to the total
amine content in the hydrothermally treated DAP catalyst than in other DAP materials.
299
3
3
00
01
302
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
303
304
Figure 2. A selected portion of the FT-IR spectra of representative cellulose nanocrystal
materials containing important carbonyl and other stretches demonstrating oxidation and
functionalization with DAP or bocDAP, as well as subsequent deprotection.
3
3
05
06
3
07
3.2 SBA-15 organocatalyst characterization
308
The SBA-15 catalysts used in this work, in particular SBA-APS, have been widely
synthesized and studied as catalysts for aldol condensations of 4-nitrobenzaldehyde and other
substrates.6,8–12,18 Their inclusion herein is primarily for comparative purposes with CNC
catalysts containing similar functional chemistry. Brunauer–Emmett–Teller (BET) surface
areas and pore volumes of SBA-15 catalysts and intermediate materials as well as silane
3
3
09
10
311
3
3
12
13
loadings are shown in Table 2. N physisorption analysis showed the expected type IV
2
314
15
isotherms (Figure S5). Surface areas and pore volumes decreased with each grafting step and
increased slightly with deprotection of the carboxylic acid-containing silane.
3
ACS Paragon Plus Environment

Page 17 of 43
ACS Catalysis
1
2
3
4
3
3
16
17
Thermogravimetric analysis and elemental analysis verified successful grafting of 3-
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
aminopropyltriethoxysilane (APS) and t-butyl 2-methyl-3-(triethoxysilyl)-propanoate
318
(Figure S6). Mass losses from ~200–300 °C are attributed to the removal of the tert-butyl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
18
3
3
19
20
protecting group. After thermal deprotection under flowing N , there was no mass lost in
2
this temperature range in a subsequent thermal ramp, suggestive of complete deprotection.
Final APS and COOH loadings were quantified with elemental analysis (nitrogen) and TGA
organic mass losses, respectively (Table 2). Silane grafting results were consistent with
studies previously reporting these methods.18 Acid (SBA-COOH-APS) and base (both
321
3
3
3
3
3
3
22
23
24
25
26
27
-1
catalysts) loadings (0.35–0.4 mmol g ) are of the same order of magnitude as acid and base
species in CNC catalysts, but a broader range of loadings and, importantly, acid:base ratios,
are found across the CNC materials.
328
Table 2. Textural properties and quantification of grafted silane species for silica-supported
3
29
organocatalysts.
a
b
c
COOH loading NH loading COOH:NH2
BET surface
Pore volume
2
Sample
c
2
-1
-
1
-1
3
-1
ratio
area (m g )
(mmol g )
(mmol g )
(cm g )
SBA-15 bare
SBA-APS
-
-
-
0.35
-
-
-
-
805
511
648
0.83
0.70
SBA-TBSP
0.36
0.69
SBA-TBSP-APS
SBA-COOH-APS
0.35
0.35
0.38
0.40
0.92
0.88
423
475
0.64
0.68
a
3
3
3
30
31
32
Quantity of protected (TBSP) or deprotected (COOH) grafted silanes determined by
thermogravimetric analysis of mass losses. Mass losses attributed to APS content were accounted
for when appropriate.
b
3
33
Primary amine content from grafted APS determined by N elemental analysis.
c
334
35
BET surface area and pore volume calculated from N physisorption data.
2
3
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
3
36
37
The synthetic methods employed were developed to achieve good dispersion of
grafted silanes, and the low loadings of aminosilanes in SBA-APS and SBA-COOH-APS are
consistent with conditions demonstrated for some degree of site-isolation.29,31,57 The relative
spatial abundance of acid and base species is not directly probed. Initial grafting of TBSP
should yield well dispersed silanes due to its relatively bulky and aprotic functional group as
well as use of the initial anhydrous stirring step in the synthesis prior to reactive hydrothermal
condensation. The dispersion of additional grafted silanes (APS) would be further improved
by the sterically bulky and randomly dispersed TBSP grafted first. The random distribution
of acidic and basic silanes on the amorphous porous surface of the SBA-15 will be discussed
further in the context of these materials’ catalytic function. As in the referenced work by
338
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
39
40
341
3
3
3
3
3
3
42
43
44
45
46
47
1
8
Brunelli et al, FTIR spectra of the deprotected bifunctional silica-supported catalyst (SBA-
COOH-APS) indicate that, despite application of synthetic methodologies aiming for for
silane dispersion, a significant portion of the carboxylic acids form carboxylates as they are
able to interact with nearby primary amines (Figure S7).
348
3
3
49
50
3.3 CNC catalysts in aldol condensation reactions
351
4-nitrobenzaldehyde (4NB), a broadly used model aldehyde reactant, as well as
furfural (furf), a biomass-derived platform molecule, were studied as aldol condensation
(AC) substrates in batch liquid phase reactions (Figure 3). Acetonitrile was used as a 1:1 v:v
co-solvent with acetone to increase the polarity of the reaction mixture, improve CNC
3
3
3
3
3
3
52
53
54
55
56
57
4
9
dispersability, and increase the effective boiling point. Acetonitrile, a polar aprotic organic
solvent, has been shown to be have minimal effect on reaction rates in 4NB-AC reactions
5
8
with heterogeneous organocatalysts. A moderate concentration (50 mM) of the aldehyde
358
substrate was used in reactions with excess acetone and 10 mol % catalyst.
ACS Paragon Plus Environment

Page 19 of 43
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
59
60
Figure 3. Aldol condensation (AC) reaction scheme of furfural (furf) or 4-nitrobenzaldehyde
(4NB) with excess acetone used for catalyst evaluation. The major products of each are the
respective aldol (1) and dehydration (2) products while small quantities of crossed aldol
adducts and side products were found.
361
3
3
62
63
3
64
The true availability of free primary amines was not precisely determined, as both
synthetic routes, coupling either a diamine (DAP) or a mono-boc protected diamine
(bocDAP), may result in fewer available primary amines relative to the theoretical quantity
(one primary amine per coupled diamine). Some crosslinked diamides may form in the
synthesis of DAP catalysts,49 and incomplete deprotection was evident in deprotected
bocDAP (dbDAP) catalysts (Figure 1, Figure 2). Catalysts are quantitatively compared by
their amine site-specific activity. Due to the ambiguity in exact primary amine availability,
site time yields (STYs) are reported rather than turnover frequencies (TOFs). To calculate
STYs, the rate of reaction in the approximately linear low-time reactant conversion (typically
0.5 or 1 h) regime, up to a conversion of 10–20 %, was normalized by the mol % catalyst
from the total diamine content. As such, STYs represent a lower bound for site-specific initial
rates as the available primary amines would be less than or equal to the theoretical content
determined by elemental analysis.
365
3
3
66
67
368
3
3
69
70
371
3
3
3
72
73
74
375
76
3
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
3
77
78
The active site loading is defined only by the amine (base) content, but the aldol
condensation is most effectively catalyzed by an acid-base cooperative mechanism (Figure
S8).12–14 Acid-base active sites may consist of an amine species with one or more proximal
acids, depending on the material. As propyl-linked primary amines are the base species
present in all materials studied herein, the relative catalytic function of reported catalysts is
determined primarily by the type, abundance, and positioning of cooperative acid partners.
The ratio of free primary amines to the theoretical value used in mol % calculations is another
key factor. Surface silanols and grafted carboxylic acids, if present, are the weakly acidic
species available in the silica-supported materials. Surface hydroxyls, sulfate half-esters, and
carboxylic acids are the acidic species present in CNC-based organocatalysts. The
effectiveness of these various acidic species in cooperative catalysis will be discussed in the
context of the catalyst series investigated.
379
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
80
81
382
3
3
3
3
3
3
83
84
85
86
87
88
389
CNC catalysts are synthesized and stored as homogeneous suspensions of individual
or small agglomerates of CNCs and are not well-defined porous or solid materials like the
mesoporous silica organocatalysts routinely studied for this aldol reaction. Reactions were
performed in suspensions of CNC particles that were partially agglomerated in the organic
co-solvent mixture. The operando CNC agglomerate form and size are not well defined, but
Mears’ criterion and Weisz-Prater parameter calculations based on reasonable assumptions
were performed to determine the potential for limiting transport effects. For both CNC and
SBA-15 catalysts, external and internal transport limitations were negligible (Table S1, Table
S2). Additionally, carboxylated CNC samples, prior to amine functionalization, have no
discernible activity for the aldol condensation under the conditions used herein (Figure S9).
3
3
90
91
392
3
3
3
3
3
3
93
94
95
96
97
98
ACS Paragon Plus Environment

Page 21 of 43
ACS Catalysis
1
2
3
4
3
4
99
00
The effect of the extent of initial removal of sulfate half-esters in CNC catalysts on
catalytic activity is shown in Figure 4. This comparison is made across catalysts that were
identically synthesized apart from their SE content. With SE content as the primary
differentiator, it was found that the partially desulfated (ds) catalyst had the greatest activity.
Sulfate half-esters are more strongly acidic and bulkier than carboxylic acid groups created
by hydroxyl oxidation. As partial sulfate removal increased activity, SEs may not be effective
participants in aldol condensation catalysis utilizing cooperative proximal acid and base sites.
Examples of aminosilica acid-base catalysts for aldol condensations in literature have shown
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
401
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4
02
03
404
4
4
4
4
4
05
06
07
08
09
1
2
that strong acids are generally less effective cooperative partners. Sulfate half-esters may
also interact too strongly with nearby primary amines and act to poison base sites, whereas
weaker and more regularly spaced carboxylic acids may not.34
4
4
10
11
Figure 4. The effect of SE content on initial aldol condensation reaction rates for reactions
of 4-nitrobenzaldehyde (4NB) and furfural (furf). The catalysts compared here (CNC-s-,
CNC-ds-, CNC-DS-cooh-dbDAP) are similar with respect to surface chemistry
characteristics except in SE content (Table 1).
412
4
4
13
14
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
4
15
16
However, near-complete removal of SEs does not further increase activity. Sulfate
half-esters are known to be critical for CNC dispersion and for maintenance of stable
suspensions. As surface charge decreases with SE removal, electrostatic repulsion decreases
and CNCs may irreversibly agglomerate, evidence of which was seen in the stark increase in
opacity and viscosity after significant desulfation. Dynamic light scattering of CNC starting
materials with varying degrees of sulfate half-ester removal as well as the catalysts compared
here indicate that significant desulfation causes particle agglomeration (Figure S10). SE
groups may not be effective for catalysis, but their presence may be critical for synthesizing
well dispersed and highly accessible CNCs. Notably, the highly desulfated (DS) material
resulted in the lowest amide coupling conversion across all CNC materials studied. Given
these disparities, highly desulfated catalysts are likely limited by agglomeration, transport,
and/or functionalization, and may not be useful for further fundamental comparisons of
surface chemistry effects on catalysis.
417
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4
18
19
420
4
4
4
4
4
4
21
22
23
24
25
26
427
4
4
28
29
CNC catalysts were also synthesized with differing extents of oxidation.
Investigations of heterogeneous organocatalytic aldol condensations have shown some base-
5
9
430
only activity, however activity is significantly increased by the presence of cooperative
weak acids. Proposed mechanisms suggest acids primarily activate relevant carbonyls, like
the ketone of acetone or the aldehyde of the, here, limiting reactant (4NB or furf) (Figure
S8). In addition to absolute differences in final carboxylic acid content, significant variation
4
4
4
4
4
4
31
32
33
34
35
36
in acid:base (COOH:NH ) ratio is noted across CNC catalysts synthesized (Table 1). The
2
cooperative nature of the catalysis suggests that acid-base site pairing is a critical metric.
5
9
Indeed, this has been studied in aminosilica acid-base catalysts. The site-specific activities
ACS Paragon Plus Environment

Page 23 of 43
ACS Catalysis
1
2
3
4
4
4
37
38
of all COOH/NH catalysts were thus compared as a function of their COOH:NH (acid:base)
2
2
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ratio in Figure 5.
439
Base-normalized activity (STY) was found to be correlated with catalyst acid:base
ratio. High SE content was shown above to reduce catalytic activity, and surface hydroxyls
are abundant (> 2.5 mmol OH/g) relative to carboxylic acids, which would reduce the strong
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4
40
41
6
0
442
correlation in activity with COOH content if these hydroxyl groups (pK 10–14 ) were
a
4
4
4
4
4
4
43
44
45
46
47
48
cooperatively effective. As such, this correlation suggests that carboxylic acids are the
primary cooperative acid species in the CNC-based catalysts. In addition, DAP-based CNC
catalysts are shown to be markedly more active than analogous dbDAP catalysts across a
range of acid:base ratios (Figure 5). Theoretically, if amidation with DAP resulted in no two-
site crosslinking and bocDAP materials were completely deprotected, the two syntheses
would produce identical materials. However, the inferior activity found in dbDAP catalysts
indicates that the issue of incomplete deprotection more significantly reduced active site
availability than diamide formation in DAP catalysts (Figure 1).
449
4
4
50
51
The dbDAP catalysts generally had greater acid:base ratios, as hydrothermal
deprotection reduced the total amine loading and liberated additional carboxylic acid groups.
However, even at a relatively low acid:base ratio (0.6), CNC-ds-COOH-DAP had greater
activity than the best dbDAP catalyst, CNC-ds-COOH-dbDAP, which had an acid:base ratio
of 1.5. The extent of diamide formation and reduction in free amine availability may be quite
minor, as the CNC crosslinking resulting from similar DAP functionalization studied in our
452
4
4
4
4
4
4
53
54
55
56
57
58
4
9
previous work was not significant. Still, both dbDAP and DAP catalysts showed a strong
positive correlation in STY with acid:base ratio (Figure 5).
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
4
59
60
The best performing CNC catalyst, CNC-ds-COOH-DAP-HT, combines all the
elements found in this study to be beneficial for catalysis. While there were no boc protecting
groups to remove, the hydrothermal treatment hydrolyzed some of the amide bonds. This
reduced total amine content, increased residual COOH loading, increased the COOH:NH2
ratio, and likely reduced the number of diamide-coupled DAP molecules. This material also
featured partially reduced sulfate half-ester content, shown earlier to be optimal among the
conditions tested. These attributes afforded a STY for the aldol condensation of 4NB of 10.
461
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4
62
63
464
4
4
65
66
-
4 -1
× 10 s .
467
68
4
Figure 5. Base site-normalized catalytic activities of COOH-containing acid-base
469
organocatalysts as a function of COOH:NH ratio in aldol condensation reactions of 4NB
2
4
4
4
70
71
72
(filled symbols) or furfural (open symbols) with acetone. Trends in STY with activity are
shown for catalyst types (DAP and dbDAP) and each aldehyde substrate. The SBA-15-
supported COOH/NH catalyst (SBA-COOH-APS) is included for comparison. The highly
2
ACS Paragon Plus Environment

Page 25 of 43
ACS Catalysis
1
2
3
4
4
4
73
74
desulfated CNC catalyst (purple) is included for reference but does not fit the trend found for
the other deprotected bocDAP catalysts (dbDAP, red).
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
475
As seen in Figure 5, catalytic activity was greater for aldol condensations of 4-
nitrobenzaldehyde than for furfural, both with acetone, for almost all catalysts studied. 4NB
is often used as a model reactant for AC reactions with acid-base organocatalysts as it is
activated by the electron-withdrawing para nitro substituent. Furfural is not similarly
activated with an EWG substituent. For both substrates, similar trends in activity with catalyst
acid:base ratio were found. The catalysts in this study have moderate acid:base ratios, mostly
between 0.5 and 1.5. An increase in acid:base ratio would indicate an increased availability
of cooperative acidic functionality for base sites. In base-normalized activity, this increase in
acid availability increases the likelihood of a substrate to be activated in the proximity of an
active base site. At high enough acid:base ratios, a maximum number of proximal carboxylic
acid species for each amine site would be reached and further increasing this ratio would not
further increase base-specific activity. The catalysts in this study do not reach this limit, but
this effect has been demonstrated in aminosilica organocatalysts.59
4
4
4
76
77
78
479
4
4
80
81
482
4
4
83
84
485
4
4
4
86
87
88
For randomly spaced and appropriately dense surface functionalization, an acid:base
ratio around one would lead to effective site-pairing and cooperativity while maximizing the
availability of total acid-base pairs. Representations of an idealized CNC catalyst including
acid-base spacing and pairing are shown in Figure 6. At very high acid:base ratios, base site-
specific activity may reach a maximum but overall catalyst active site loading would be
necessarily lower. The highly desulfated catalyst discussed earlier does not fit in the trends
that are apparent across the rest of the materials studied (Figure 5). This is additional evidence
489
4
4
90
91
492
4
4
93
94
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
4
95
96
that excessive desulfation introduces confounding catalyst limitations, such as
agglomeration, that reduce activity and prevent fundamental investigation of catalyst activity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
497
498
Figure 6. Spatial positioning and structure of carboxylic acid (red) and amine (blue) surface
species for the two crystalline faces present in the cellulose I훽 structure of wood-derived
cellulose nanocrystals. Oxygen atoms not in carboxylic acid species are shown in pink, and
carbon atoms are gray. In all depictions, hydrogen atoms are omitted to reduce clutter, and
SE content is omitted for simplicity. Theoretical DAP-functionalized CNC surfaces with
4
5
99
00
501
5
5
5
02
03
04
complete initial oxidation of C6 hydroxyls, 1:1 COOH:NH ratios, and optimal checkerboard
2
acid-base arrangements are represented. Crystallite widths were chosen to approximate a
typical CNC cross-sectional size. Arrows indicate the truncated z-axis of the cellulose
crystallite. (a) – (c) refer to the (010) surfaces and (d) – (f) to the (100) surfaces. (a) and (d)
505
5
5
5
5
06
07
08
09
Top-down views of the (010) and (100) surfaces, respectively. 훿 (010) and 훿 (100) are
1
3
representative NH – COOH distances along a single cellulose chain (intra-chain), and 훿
2
2
(010) and 훿 (100) are representative inter-chain spacings. Average values for these distances
4
ACS Paragon Plus Environment

Page 27 of 43
ACS Catalysis
1
2
3
4
5
10
with pseudo-randomly oriented acid and base surface species are: 훿 = 10.6 ± 2.2 Å, 훿 =
1
2
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
511
8.4 ± 1.7 Å, 훿 = 11.0 ± 2.3 Å, and 훿 = 9.3 ± 1.8 Å. (b) and (d) Z-axis (into the page)
3 4
5
5
12
13
views of the first three layers of the catalyst surface with end functionalities ignored. (c) and
(e) Angled surface perspectives of the two catalyst surfaces. (g) A full crystallite cross-
section (z-truncated, omitting chiral twist) of an unfunctionalized CNC surface.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
514
515
3.4 Product selectivity and comparisons between CNC vs. SBA-15 organocatalysts
The CNC-based organocatalysts developed in this work were compared to SBA-15-
supported aminosilica catalysts with analogous surface chemistry: propyl-linked primary
amines and a cooperative weak acid, either surface silanols (SBA-APS) or co-grafted silanes
with a carboxylic acid moiety (SBA-COOH-APS). The SBA-APS catalyst used is more
similar to the current state-of-the-art within this material class,11,61 but the SBA-COOH-APS
catalyst was also included as a more direct analogue in terms of surface chemistry.18
Under the reaction conditions used in this study, the best-performing CNC catalyst
had slightly greater site-specific activity for the aldol condensations of 4NB and furfural than
the well-optimized SBA-APS catalyst (Figure 7). As expected from previously reported
5
5
16
17
518
5
5
5
5
5
5
19
20
21
22
23
24
1
8
525
findings, SBA-COOH-APS demonstrated significantly reduced activity compared to SBA-
5
5
26
27
APS. It has been shown that surface silanols are particularly effective cooperative weak acids
8
in simple aminosilane-functionalized silica-supported organocatalysts. This is due to their
1
8
528
low acid strength (pK ~ 4.5–8.5 ), which prevents strong poisoning interactions with
a
5
5
5
29
30
31
grafted amines, and to their high relative availability, leading to high acid:base ratios. For
isolated grafted aminosilanes on silicate surfaces, they are effectively surrounded by
cooperative weak acid sites, maximizing their site-specific activities. However, the
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 43
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
5
32
33
referenced work that investigated co-grafted carboxylic acid silanes18 concluded that
ineffective dispersion and site pairing as well as strong acid-base interactions led to the low
activity measured.
534
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
5
35
36
The CNC catalysts studied could have suffered from similar limitations as they
contain nearly identical acid and base moieties to SBA-COOH-APS. However, base site-
specific activities greater than those of site-isolated aminosilanes grafted on SBA-15 (SBA-
APS) with plentiful cooperative silanols were achieved. This suggests that effective
bifunctional acid-base cooperativity between a carboxylic acid and a primary amine may be
possible, but that the synthetic approach and material characteristics of the SBA-15 catalysts
hinder this outcome. Moreover, the CNC acid-base catalysts do not appear to have reached
their maximum activity, as the data in Figure 5 suggest that further increasing the acid:base
ratio could yield additional improvements to site-specific activity. This effective COOH–
537
5
5
5
5
5
5
38
39
40
41
42
43
544
NH acid-base cooperativity counters the findings of previous studies of aminosilica catalysts
2
concluding that silanols are inherently superior cooperative acids than carboxylic acids.18,34
5
5
45
46
This conclusion was also based on AC reactions of the model substrate 4NB and did not
consider product selectivity, only catalytic activity calculated from initial rates.
547
5
5
5
5
5
5
48
49
50
51
52
53
The catalysts were also evaluated for their selectivity to the AC dehydration product
(2, Figure 3) after the 6 h reaction period. Simple aminosilica organocatalysts like SBA-APS
typically show low selectivity to the dehydration product, with the major product being the
aldol product 1. CNC catalysts investigated showed high selectivity, typically >80 %, toward
the dehydration product in reactions of 4NB and furfural across all materials (Figure 7, Table
S3), significantly greater than the selectivities of the SBA-15 catalysts in reactions of 4NB.
SBA-APS contains only weakly acidic silanols, while the CNC acid-base catalysts
554
ACS Paragon Plus Environment

Page 29 of 43
ACS Catalysis
1
2
3
4
5
5
55
56
synthesized have a significant surface loading of stronger carboxylic acids (pK ~4). The
a
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
availability of these stronger acids likely facilitates the acid-catalyzed dehydration of 1 to 2,
leading to the selectivity results observed. In CNC catalysts with poorer activities, with which
reactant conversions after 6 h were relatively low, selectivities to 2 remained high.
Selectivities were fairly consistent across all CNC materials, despite the wide range of
reactant conversions in the 6 h product mixtures for various catalysts. This indicates a rapid
dehydration of 1 to 2, even at low concentrations of the aldol product 1. Time-resolved
reaction selectivity was determined from scaled reactions with the optimized CNC catalyst
(CNC-ds-COOH-DAP-HT) and SBA-APS (Figure S11). Selectivity to 2 was approximately
constant over the 6 h reaction for both materials, indicating that the reaction selectivity is an
intrinsic property of the catalyst and reaction environment and not of the reaction time or
procession. Even at low initial conversions, the CNC catalyst achieved close to 90 %
selectivity, supporting the claim of rapid dehydration.
557
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
5
58
59
560
5
5
5
5
5
5
61
62
63
64
65
66
567
ACS Paragon Plus Environment