Synthesis of cellulose adipate derivatives

Article abstract of DOI:10.1021/bm101448f

Cellulose esters containing adipates and other ester groups are synthesized by the reaction of commercially available cellulose esters in solution with the benzyl monoester of adipoyl chloride. The products, cellulose adipate esters in which the distal en

Full text of DOI:10.1021/bm101448f

ARTICLE
pubs.acs.org/Biomac
Synthesis of Cellulose Adipate Derivatives
Nilanjana Kar, Haoyu Liu, and Kevin J. Edgar*
Macromolecules and Interfaces Institute, Virginia Polytechnic Institute, Blacksburg, Virginia 24061, United States
S Supporting Information
b
ABSTRACT:
Cellulose esters containing adipates and other ester groups are synthesized by the reaction of commercially available cellulose esters
in solution with the benzyl monoester of adipoyl chloride. The products, cellulose adipate esters in which the distal end of the adipate
moiety is a benzyl ester, were easily converted to cellulose adipate derivatives by Pd-catalyzed hydrogenation. These cellulose
adipate derivatives are promising biopolymers for drug delivery and other applications in which water-dispersion or swelling are
desired.
’ INTRODUCTION
to be a promising polymer for drug delivery formulations that
significantly enhance drug solubility by entrapping the drug in
metastable amorphous form dispersed in the polysaccharide
matrix. Another pertinent example is carboxymethylcellulose
acetate butyrate (CMCAB), which has been recently shown11,12
to combine several desirable drug delivery functions; pH-controlled
release, zero-order release (which is of great value since it can permit
constant blood concentration of the drug over long time periods,
minimizing side effects and reducing dosing frequency), and
solubility enhancement by the amorphous dispersion method.13À17
CMCAB and HPMCAS function by swelling in the neutral pH of
the small intestine and colon because of ionization of the carboxyl
groups; this swelling promotes drug release. The carboxyl groups of
both cellulose derivatives and the hydrophobic substituents, parti-
cularly of CMCAB, promote miscibility of hydrophobic drugs in the
polysaccharide matrix, helping to stabilize amorphous dispersions of
otherwise poorly water-soluble drugs.15 The kinetic stability of the
thermodynamically metastable amorphous drug in the polymer
matrix, for up to several years of storage at relatively high tempera-
tures and humidity, is a critical performance feature of the amor-
phous dispersion formulations that are showing such promise for
enhancing bioavailability of poorly water-soluble drugs. (There is at
least one commercial amorphous formulation of the HIV drugs
lopinavir and the CYP3A4 inhibitor ritonavir.18) Such kinetic stability
should be enhanced both by the high Tg of cellulose derivatives and by
the specific and hydrophobic interactions provided by hydrophobic
cellulose derivatives that contain ionizable carboxyl groups.
The synthesis of water-dispersible derivatives of cellulose has
long been a difficult problem of significant scientific and com-
mercial importance. Synthesis of water-soluble cellulose deriva-
tives is relatively well-understood, for example, by acylation,
followed by extensive back-hydrolysis,1 or by reaction of cellulose
with various alkylating agents under base catalysis to make
derivatives like methylcellulose,2,3 carboxymethylcellulose,4
hydroxyethylcellulose,5 or ethylcellulose.6,7 These methods take
advantage of the native hydrophilicity of cellulose and the fact
that a low DS of nearly any substituent can render cellulose
water-soluble by disrupting hydrogen bonding and crystallinity.
For many applications, the ability to swell or disperse in water is
far more desirable than water solubility. For example, in coatings
applications, the low viscosity of aqueous dispersions permits the
formulation of high-solids coatings, reducing drying time and
transportation costs by minimizing the water content of the
coating dispersion. One of the most straightforward ways to
make an aqueous dispersion that can then coalesce into a film is
to incorporate carboxyl groups into the molecule; these can be
deprotonated by volatile amines, enhancing the dispersibility of
the polymer.8 Upon drying, the amine evaporates, promoting the
formation of a continuous film. Polysaccharides that are funda-
mentally hydrophobic but contain carboxyl groups that enable
swelling or dissolution at neutral pH are also highly useful for
drug delivery applications.9 Recent studies have shown that not
only are such polysaccharide derivatives useful for traditional
drug delivery applications such as pH-controlled release to avoid
contact between drug and stomach,9 but they also provide more
valuable functional benefits. Recently, for example, hydroxypro-
pylmethylcellulose acetate succinate (HPMCAS)10 has proved
Received: December 1, 2010
Revised:
March 2, 2011
Published: March 07, 2011
r
2011 American Chemical Society
1106
dx.doi.org/10.1021/bm101448f Biomacromolecules 2011, 12, 1106–1115
|

Biomacromolecules
ARTICLE
Scheme 1. Proposed CAB-Su Hydrolysis Mechanism
Scheme 2. Predicted CE-Adipate Hydrolytic Stability
homopolymerization and has most commonly been used to
synthesize polyanhydrides29 and polyesters.30 Given its reactivity,
we hoped that adipic anhydride31 would react with the poorly
nucleophilic cellulose in solution, providing that we could find an
appropriate solvent and catalyst system. We report herein the
unexpected difficulties we encountered in direct reaction of
cellulose with adipic anhydride and how those were overcome
by employing monofunctional adipate reagents to develop the first
successful synthesis of cellulose adipate esters, which will permit
the synthesis of a broad range of cellulose adipate alkanoates.
Unfortunately, there are a limited number of methods avail-
able for synthesis of carboxyl-containing polysaccharides and, in
particular, carboxyl-containing cellulose derivatives. The most
well-known synthesis of carboxyl-containing cellulose derivatives
is that of the commercial aqueous thickener carboxymethylcel-
lulose. Typically, cellulose is reacted with chloroacetic acid in
water in the presence of sodium hydroxide as base. Conversion of
carboxymethylcellulose into hydrophobic ester derivatives like
acetate, propionate, and butyrate has been reported.19,20 Whereas
the resulting carboxymethylcellulose esters are very interesting
materials for coatings and drug delivery applications,21 there must
be concern with acid-catalyzed esterification of a polymer that
contains both carboxyl and hydroxyl groups; the potential for
cross-linking by esterification must always be considered. Another
method is the reaction of cellulose or a cellulose derivative with a
cyclic anhydride, usually employing a basic catalyst such as
pyridine or triethylamine. In this way, carboxyl-substituted deri-
vatives of cellulose such as cellulose acetate phthalate22 and
cellulose acetate butyrate succinate23 have been synthesized. The
known carboxyl-substituted cellulose derivatives prepared by this
cyclic anhydride ring-opening chemistry suffer from limited stabi-
lity in aqueous systems, which is of course a drawback for polymers
designed to work in aqueous media. The stability of cellulose
acetate butyrate succinate in neutral to alkaline media, for example,
islimited.24 The authors of that study speculatedthat the particular
instability of derivatives like succinate arises from their ability to
catalyze their own hydrolysis (Scheme 1).
Cellulose adipate derivatives may be expected to possess super-
ior properties to those of the cellulose phthalates and succinates
synthesized in previous work. They are likely to be less prone to
autocatalyzed hydrolysis because this would require an unfavorable .
seven-membered ring transition state (Scheme 2). They will be
more hydrophobic than the corresponding succinate or phthalate
derivatives and thus would be expected to be more compatible with
hydrophobic drugs. To date, there have been no reports of the
synthesis of well-characterized cellulose adipates. Polysaccharide
materials have been subjected to surface treatments with adipic
acid to cross-link and strengthen those materials,25À28 but the
products were not fully characterized, and no discrete cellulose
derivatives were described.
’ EXPERIMENTAL SECTION
Materials. The high-purity hardwood pulp used in these experi-
ments was Sulfatate HJ from Rayonier. Cellulose acetate propionate
(CAP-504-0.2), cellulose acetate butyrate (CAB-553-0.4), and cellulose
acetate (CA-320S) were from Eastman Chemical. Methyl ethyl ketone
(MEK) was dried by refluxing over potassium carbonate. Other pur-
chased reagents were used as received. Adipic anhydride31 was synthe-
sized by following previously reported procedures. 10% Pd/C and
20% Pd(OH)2/C were purchased from Sigma Aldrich. Adipic acid,
p-toluenesulfonic acid (PTSA), 1,3-dimethyl-2-imidazolidinone (DMI),
oxalyl chloride, and triethylamine (Et3N) were purchased from ACROS
Organics. Benzyl alcohol, toluene, N,N-dimethylformamide (DMF), N,
N-dimethylacetamide (DMAc), dichloromethane, lithium chloride, and
glacial acetic acid were purchased from Fisher Scientific. DMAc and
DMI were dried over 4 Å molecular sieves. Anhydrous tetrahydrofuran
was purchased from Sigma Aldrich. All other reagents were synthesized
by procedures described later in the Experimental Section.
Measurements. DSC analyses were obtained using a TA Instru-
ments Q100 apparatus. Neat powders (1À3 mg) were loaded in
aluminum hermetic pans. Dry N2 was used as the purge gas at
50 mL/min. Samples were equilibrated at À40 °C and then ramped
with a heat/cool/heat procedure between À40 and 160 °C at a heating
rate of 10 °C/min and a cooling rate of 5 °C/min. All tests were
performed in triplicate. Tg values were recorded as the step-change
inflection point from second heat scans. IR spectra were obtained on a
Nicolet 8700 instrument. Proton NMR spectra were acquired on an
INOVA 400 spectrometer operating at 400 MHz. The sample tube size
was 5 mm, and the sample concentrations were ca. 10 mg/mL in CDCl3
or DMSO-d6. CP/MAS 13C NMR was carried out using a Bruker Avance
II 300 instrument using 2048 scans at 27 °C. Elemental analyses were
carried out by Micro-Analysis. Size exclusion chromatography (SEC)
was used to determine the molecular weights of the polymers at 50 °C in
DMF with 0.05 M LiBr at 1 mL minÀ1 flow rate. The DMF SEC system
was equipped with a Waters 717 plus autosampler, a Waters 1525 HPLC
pump, two Waters Styragel HR5E (DMF) columns, and a Waters 2414
differential refractive index detector. Number-average molecular weights
relative to polystyrene standards are reported. Dynamic light scattering
(DLS) was used to confirm the absence of polymer aggregates in
solution. Solutions were prepared at a concentration of 1.0 mg/mL in
Adipic anhydride would be the directly analogous reagent to
consider for ring-opening reaction with cellulose to produce
cellulose adipates. Adipic anhydride is highly reactive because of
its relatively strained seven-membered ring structure. It is prone to
1107
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115

Biomacromolecules
ARTICLE
pure DMF and DMF with 0.01 M LiBr and analyzed in a quartz cuvette
at 20.0 °C. DLS measurements were performed on a Malvern ZetaSizer
Nano Series Nano-ZS instrument using Dispersion Technology Soft-
ware (DTS) version 6.01 at a wavelength of 633 nm using a 4.0 mW,
solid state HeÀNe laser at a backscattering angle of 173°. The experi-
ments were performed at a temperature of 20 °C in triplicate. Solubility
testing on cellulose ester samples was performed by adding 10 mg of
sample to a glass vial, then adding 2 mL of solvent. The mixture was
subjected to vortex mixing for 5À10 min at room temperature; then,
solubility was judged by visual examination.
Reaction of Cellulose in DMAc/LiCl Solution with Adipic
Anhydride (Scheme 3). This procedure is typical for entries listed in
Table 1. Cellulose (8.00 g, 49.3 mmol) was dissolved in DMAc (300 mL)
and lithium chloride (15 g) by the procedure previously reported.32 To
this solution, at 80 °C under nitrogen (entry 1, Table 1) was added
dropwise a solution of adipic anhydride (6.31 g, 49.3 mmol) in 20 mL of
DMAc (entry 1, Table 1). After 45 min, it was observed that the solution
gelled, and the gel broke up into small, translucent balls (ca. 1 to 2 cm in
diameter) with mechanical stirring. The product was isolated by adding
the reaction mixture to methanol, filtration of the gel-like product, and
then extensive washing of the product with methanol, then with water.
The product was insoluble in all solvents tried, including DMSO and
chloroform. Analysis of the dried product by infrared spectroscopy and by
solid-phase13Cspectroscopy revealed that itwasa cellulose adipate, which
was cross-linked. Solid state 13C NMR (Figure 1) showed adipate peaks,
but the IR spectrum (Figure 2) showedno prominent COOH absorption,
indicating the lack of pendant carboxylic acid groups and affirming the
presence of ester cross-links. Solid-state CP/MAS 13C (ppm): 173.5
(CdO), 104.6 (C-1), 83.1 (C-4), 74.8(C-5, C-2, C-3), 62.7 (C-6), 33.8À
38.5 (COCH2CH2CH2CH2CO of adipate), 21.7À24.6 (COCH2CH2
CH2CH2CO of adipate). IR (cmÀ1): 1812, 1742 (anhydride doublet).
Synthesis of Monobenzyl Adipate (Scheme 5). Prepared by
the procedure of English et al.,32 as follows. Adipic acid (73 g, 0.5 mol),
benzyl alcohol (81 g, 0.75 mol), PTSA (0.95 g, 5 mmol), and toluene
(400 mL) were combined in a flask equipped with DeanÀStark trap and
heated at reflux until the theoretical amount of H2O (13.5 mL, 0.75 mol)
was obtained. The solution was then cooled, 300 mL of H2O was added,
and the pH was adjusted to 8 with 6 N NaOH. The aqueous layer was
separated and washed with ether (2 Â 100 mL), 200 mL of fresh ether
was added, and the pH was adjusted to 2.0 with 6 N HCl. The ether layer
was separated, dried (Na2SO4), and concentrated to yield 47 g of
Scheme 3. Reaction of Cellulose with Adipic Anhydride
Table 1. Synthesis of Cellulose Adipate under Different
Conditions Using Adipic Anhydride
1
monobenzyl adipate (40%) as a colorless oil: H NMR (CDCl3) 1.68
(m, 4 H), 2.36 (m, 4 H), 5.09 (s, 2 H), and 7.32 (m, 5 H).
Synthesis of Monobenzyl Adipoyl Chloride (Scheme 5).
Prepared by the procedure of Abell et al.,33 as follows. A solution of
monobenzyl adipate (20 g, 0.08 mol) and DMF (1 drop) in dichlor-
omethane was cooled to 0 °C, and oxalyl chloride (57.15 g, 0.45 mols)
was slowly added. After 30 min at 15 °C, the solvent was removed under
entry
adipic anhydride (equiv)
temperature (°C)
remarks
1
2
3
1
80
40
60
cross-linked
no reaction
cross-linked
1
3.5
Figure 1. CP/MAS 13C NMR spectrum of cross-linked cellulose adipate.
1108
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115

Biomacromolecules
ARTICLE
Figure 2. IR spectrum of cross-linked cellulose adipate.
reduced pressure. Toluene (200 mL) was added to the resultant oil and
again the solvent was removed to yield the acid chloride as an oil that was
not purified further. 1H NMR (CDCl3): 1.73 (m, 4 H), 2.39 (t, 2 H),
2.90 (t, 2 H), 5.12 (s, 2 H), 7.32 (m, 5 H).
dissolved in N,N- dimethylimidazolidinone (DMI, 10 mL), and the
solution was heated to 100 °C with stirring under nitrogen. Triethyla-
mine (0.63 mL, 4.59 mmol, 1.1 equiv) was added all at once; then,
monobenzyl adipoyl chloride (2.12 g, 8.36 mmol, 2 equiv) was added.
The solution was stirred at 100 °C for 20 h. The reaction mixture was
filtered to remove the needle-shaped crystals of triethylammonium
chloride. The filtrate was added dropwise to water at room temperature
in a high shear mixer. The precipitate was redissolved in isopropanol and
precipitated with water. The product, benzyl cellulose acetate adipate,
was washed with water and vacuum-dried. 1H NMR (CDCl3): 1.66 (br s,
COCH2CH2CH2CH2CO of adipate), 1.94À2.09 (m, COCH3 of
acetate), 2.37 (br s, COCH2CH2CH2CH2CO of adipate), 3.25À5.31
Example Procedure for the Reaction of CAP with Mono-
benzyl Adipoyl Chloride (Scheme 6). CAP-504-0.2 (entry 1,
Table 2) (1 g, 3.52 mmol) was dissolved in MEK (10 mL), and the
solution was heated to 60 °C with stirring under nitrogen. Et3N (0.53 mL,
3.87 mmol, 1.1 equiv) was added all at once; then, monobenzyl adipoyl
chloride (0.89 g, 3.52 mmol, 1 equiv) was added. An immediate
precipitate (presumed to be triethylammonium chloride) was observed.
The solution was stirred at 60 °C for 20 h. The reaction mixture was
filtered and then added dropwise to isopropyl alcohol at room tempera-
ture with stirring. The precipitate was collected by filtration and washed
with water. It was redissolved in chloroform and precipitated with hexane.
The product was washed with hexane and vacuum-dried at 50 °C.
A similar procedure was followed for the reaction of CAB with
monobenzyl adipoyl chloride.
1
(cellulose backbone), 5.10 (CH2C6H5), 7.33 (CH2C6H5). DS by H
NMR: adipate 0.7, acetate 1.8. Yield: 82%.
Procedure for the Reaction of Cellulose with Monobenzyl
Adipoyl Chloride. Cellulose (0.63 g, 3.88 mmol) was dissolved in
DMAC/LiCl by a previously reported procedure.34 To this solution at
60 °C was added triethylamine (0.59 mL, 1.1 equiv). Then, the
monobenzyl adipoyl chloride (3 g, 11.64 mmol, 3 equiv) was added
dropwise at 60 °C. The reaction mixture was stirred for 20 h at that
temperature and then cooled to room temperature. It was filtered and
then added to isopropyl alcohol to precipitate the product. The product
benzyl cellulose adipate was washed with water, and then was vacuum-
dried at 40 °C to afford 1.83 g (78% yield) of white powder. 1H NMR
(DMSO-d6): 1.47 (br s, COCH2CH2CH2CH2CO of adipate), 2.37
(br s, COCH2CH2CH2CH2CO of adipate), 3.25À5.31 (cellulose back-
bone), 5.10 (CH2C6H5), 7.33 (CH2C6H5). DS by 1H NMR: adipate 2.0.
Yield: 78%.
Benzyl Cellulose Acetate Adipate Propionate. 1H NMR
(CDCl3): 1.02À1.20 (m, COCH2CH3 of propionate), 1.66 (broad s,
COCH2CH2CH2CH2CO of adipate), 2.16À2.35 (m, COCH2CH3 of
propionate, COCH3 of acetate and COCH2CH2CH2CH2CO of
adipate), 3.25À5.24 (cellulose backbone), 5.10 (CH2C6H5), 7.33
1
(CH2C6H5). Degree of substitution (DS) by H NMR: adipate 0.33,
propionate 2.09, acetate 0.04. Yield: 58%.
Benzyl Cellulose Acetate Adipate Butyrate. 1H NMR
(CDCl3): 0.89À0.98 (m, COCH2CH2CH3 of butyrate), 1.54À1.64
(m, COCH2CH2CH3 of butyrate, COCH2CH2CH2CH2CO of adipate),
2.14À2.31 (m, COCH2CH2CH3 of butyrate, COCH3 of acetate and
COCH2CH2CH2CH2CO of adipate), 3.25À5.31 (cellulose backbone),
5.10 (CH2C6H5), 7.33 (CH2C6H5). DS by 1H NMR: adipate 0.5,
butyrate 1.99, acetate 0.14. Yield: 60%.
Example Procedure for the Pd/C Hydrogenation of the
Benzyl Cellulose Ester. Benzyl cellulose acetate adipate propionate
(200 mg) was dissolved in THF and to this solution was added 10%
palladium on carbon (100 mg). A hydrogen balloon was attached to the
flask, and the solution was stirred overnight under a hydrogen atmosphere
Procedure for the Reaction of CA-320S with Monobenzyl
Adipoyl Chloride. CA-320S (entry 4, Table 2) (1 g, 4.18 mmol) was
1109
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115

Biomacromolecules
ARTICLE
Table 2. Synthesis of Benzyl Cellulose Adipate Derivatives Using Monobenzyl Adipoyl Chloride (MAC)a
DS
entry
substrate
MAC (equiv/AGU)
solvent
MEK
temp (°C)
solubility
Ad
other
total
1
2
3
4
5
CAP-504-0.2
CAP- 504-0.2
CAB-553-0.4
CA-320-S
1
3
1
2
3
60
60
0.33
0.50
0.25
0.70
2.00
Pr 2.09 Ac 0.04
Pr 2.09 Ac 0.04
Bu 1.99 Ac 0.14
Ac 1.80
2.46
2.63
2.38
2.50
2.00
THF, CHCl3, DMSO
THF, CHCl3, DMSO
THF, CHCl3, DMSO
THF, CHCl3, DMSO
sparingly soluble in DMSO
MEK
MEK
60
DMI
100
cellulose
DMAc/LiCl
60
none
a Ac = acetate, Bu = butyrate, Pr = propionate. CAP-504-0.2 is commercial cellulose acetate propionate from Eastman (DS(Pr) = 2.09, DS(Ac) = 0.04).
CAB-553-0.4 is commercial cellulose acetate butyrate from Eastman (DS(Bu) = 1.99, DS(Ac) = 0.14). CA-320S is commercial cellulose acetate from
Eastman (DS(Ac) = 1.80).
at room temperature. The mixture was filtered through Celite and
evaporated to afford the product. The product was dissolved in chloro-
form and reprecipitated into hexane. The product (150 mg, 82% yield)
was soluble in DMSO and insoluble in water.
Scheme 4. Reaction of Cellulose with Poly(adipic
anhydride)
1
Cellulose Acetate Adipate Propionate. H NMR (CDCl3):
1.02À1.20 (m, COCH2CH3 of propionate), 1.66 (broad s, COCH2-
CH2CH2CH2COofadipate), 2.16À2.35 (m, COCH2CH3 ofpropionate,
COCH3 of acetate and COCH2CH2CH2CH2CO of adipate), 3.25À5.24
1
(cellulose backbone). DS by H NMR: adipate 0.33, propionate 2.09,
acetate 0.04. 13C NMR (DMSO-d6): 172.0À175.0 (CdO), 99.6 (C-1),
75.1(C-4), 72.8 (C-3), 72.1 (C-2, C-5), 62.8(C-6), 33.2À34.0(COCH2-
CH2CH2CH2CO of adipate), 26.8À27.6 (COCH2CH3 of propionate),
24.0À24.8 (COCH2CH2CH2CH2CO of adipate), 23.1 (COCH3 of
acetate), 9.1 (COCH2CH3 of propionate). Yield: 82%. Elemental anal-
ysis: Calculated C, 53.24%; H, 6.53%; O, 40.23%; Found C, 53.10%; H,
6.61%; O, 40.22%.
’ RESULTS AND DISCUSSION
We initially explored the solution phase reaction of cellulose
with adipic anhydride in DMAc/LiCl (Scheme 3), which is one
of the best understood and most broadly useful solvent systems
for reactions in which cellulose is a nucleophile.35,36 We found in
previous studies34 that cellulose reacts smoothly in DMAC/LiCl
with carboxylic anhydrides (cyclic or otherwise) at 80 °C or
higher without the need for added catalyst. Therefore, we added
adipic anhydride dropwise to cellulose in DMAC/LiCl solution
at 80 °C (entry 1, Table 1). We were surprised to observe that the
solution gelled almost immediately upon commencement of
adipic anhydride addition. The gelled product was insoluble in
all solvents tried, including DMSO and chloroform. Whereas it is
possible that the observed gelation could have been caused by
inherent poor solubility of cellulose adipate, this would not be
predicted from previous observations of carboxyalkyl cellulose
ester solubility37 and was not supported by subsequent observa-
tions of the solubility of cellulose adipates prepared in other ways
(vide infra). Analysis of the dried product by infrared spectros-
copy and solid-phase 13C spectroscopy revealed that it was a
cellulose adipate, which was cross-linked. Solid-state 13C NMR
(Figure 1) showed adipate peaks at δ 21.7À24.6 and 33.8À38.5,
and the IR (Figure 2) spectrum showed no prominent COOH
absorption, indicating the lack of pendant carboxylic acid groups.
Interestingly, a doublet characteristic of anhydrides (in this
case due to poly(adipic anhydride))36 was seen at 1812 and
1742 cmÀ1 in the infrared spectrum.
Cellulose Acetate Adipate Butyrate. 1H NMR (CDCl3):
0.89À0.98 (m, COCH2CH2CH3 of butyrate), 1.54À1.64 (m,
COCH2CH2CH3 of butyrate, COCH2CH2CH2CH2CO of adipate),
2.14À2.31 (m, COCH2CH2CH3 of butyrate, COCH3 of acetate and
COCH2CH2CH2CH2CO of adipate), 3.25À5.31 (cellulose backbone).
1
DS by H NMR: adipate 0.5, butyrate 1.99, acetate 0.14. 13C NMR
(DMSO-d6): 171.2À175.0 (CdO), 99.9 (C-1), 76.5 (C-4), 72.0À73.0
(C-2, C-3, C-5), 63.1 (C-6), 35.9 (COCH2CH2CH3 of butyrate), 34.1
(COCH2CH2CH2CH2CO of adipate), 24.8 (COCH2CH2CH2CH2
CO of adipate), 21.0 (COCH3 of acetate), 18.2À18.6 (COCH2CH2
CH3 of butyrate), 13.9 (COCH2CH2CH3 of butyrate). Yield: 85%.
Elemental analysis: Calculated C, 55.69%; H, 7.14%; O, 37.17%; Found
C, 55.59%; H, 7.11%; O, 36.89%.
Cellulose Acetate Adipate. 1H NMR (DMSO-d6): 1.47 (s,
COCH2CH2CH2CH2CO of adipate), 1.84À2.05 (m, COCH3 of acet-
ate), 2.18 (s, COCH2CH2CH2CH2CO of adipate), 2.95À5.15 (cellulose
backbone). DS by 1H NMR: adipate 0.7, acetate 1.8. 13C NMR (DMSO-
d6): 168.0À175.0 (CdO), 99.5 (C-1), 76.4 (C-4), 70.3À73.7 (C-2, C-3,
C-5), 62.8 (C-6), 33.2À34.0 (COCH2CH2CH2CH2CO of adipate),
24.0À24.5 (COCH2CH2CH2CH2CO of adipate), 20.3-21.3 (COCH3
of acetate). Yield: 88%. Elemental analysis: Calculated C, 50.61%; H,
5.87%; O, 43.52%; Found C, 46.73%; H, 6.07%; O, 45.66%.
Example Procedure for the Pd(OH)2/C Hydrogenation of
the Benzyl Cellulose Ester. To a solution of 200 mg benzyl cellulose
acetate adipate propionate dissolved in 30 mL of anhydrous THF/acetic
acid (1:1, v/v), 100 mg palladium hydroxide, 20 wt % Pd (dry basis) on
carbon was added. A hydrogen balloon was attached to the flask, and the
solution was stirred overnight under hydrogen atmosphere at room
temperature. The mixture was filtered through Celite and concentrated.
The product was dissolved in dichloromethane and precipitated into
hexane. The precipitate was collected and dried under vacuum. Yield:
130 mg (71%).
The cross-linking observed is extraordinarily rapid; further-
more, no such cross-linking is observed when employing succinic
anhydride as the electrophile under otherwise identical condi-
tions.37 Therefore, it seems extremely unlikely that the rapid
gelation observed could be caused by condensation between the
pendant carboxyl group of adipate groups attached to cellulose
and free hydroxyl groups on cellulose (nor would such ester
1110
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115

Biomacromolecules
ARTICLE
Scheme 5. Monobenzyl Adipoyl Chloride Synthesis
Scheme 6. Synthesis of Cellulose Adipates via Acid Chloride
If our conclusions about the mechanism for the cross-linking
observed with adipic anhydride were sound, then a sure-fire way
to avoid cross-linking would be to replace adipic anhydride with a
monofunctional equivalent. Fortunately for us, there were reports
in the literature of the monobenzyl ester of adipic acid,32 and the
conversion of that ester to the monobenzyl ester monoacid
chloride33 (benzyl 6-chloro-6-oxohexanoate, Scheme 5). We knew
from prior reports3,41,42 that the removal of benzyl ether groups
from cellulose by hydrogenation was feasible and were hopeful
that our adipate benzyl ester groups would be at least as easily
removed by hydrogenation. We first examined the reaction of
monobenzyl adipoyl chloride with commercial cellulose esters of
relatively low degree of substitution; this permitted us to work in
solvents of moderate polarity. Reaction of monobenzyl adipoyl
chloride with CAP-504-0.2 (DS (propionyl) 2.09, DS(acetyl)
0.04) (entry 1, Table 2) in MEK at 60 °C for 20 h cleanly afforded
benzyl cellulose acetate adipate propionate, with no evidence of
gelation or cross-linking. The benzyl adipate substituent was then
hydrogenated to remove the benzyl groups and deprotect the
adipate groups; again, the reaction went smoothly and no cross-
linking was observed, affording cellulose acetate adipate propio-
nate (Scheme 6; note that no regiospecificity is implied by the
structures in Scheme 6; we have no conclusive information at this
time about the regiospecificity of the reaction). We found that
hydrogenation using Pd/C was always successful but sluggish on
larger scale (multigram) reactions. The use of Pd(OH)2/C gave
more rapid hydrogenations on the gram scale and eliminated the
occasional necessity for multiple hydrogenations to attain com-
plete debenzylation.
cross-links be consistent with the IR and 13C analyses of the
product). Rather, we postulate that the cross-linking is being
caused by reaction between cellulose and poly(adipic anhydride)
(Scheme 4) in which each poly(adipic anhydride) chain reacts on
average with more than one cellulose molecule. We have
observed that even freshly prepared adipic anhydride contains
1
a measurable amount of poly(adipic anhydride) by H NMR
spectroscopy (poly(adipic anhydride) gave signals at 1.75 and 2.5
ppm due to ÀCH2CH2À and CH2ÀCOÀOÀCOÀCH2 pro-
tons, respectively, and adipic anhydride signals appeared at 2.8
and 2.0 ppm), and that this amount grows over the course of days
as the adipic anhydride is stored, even if refrigerated. This
postulate is consistent with the observed anhydride carbonyl
absorbances in the IR spectrum (1812, 1742 cmÀ1). These
poly(anhydride) cross-links would not be expected to hydrolyze
rapidly (even though thermodynamically labile) because of the
hydrophobicity of the adipate group and the poor solubility of
these high-molecular-weight cross-linked species.38,39
We made several attempts to circumvent the apparent cross-
linking problem by adjusting reaction conditions. Table 1 shows
the results of several of these attempts. Lower temperature
(40 °C, entry 2) condensation was attempted in an effort to
enhance reaction selectivity and minimize cross-linking, but at
that low temperature, no desired esterification was observed. We
thought that by increasing the number of equivalents of adipic
anhydride (entry 3) and employing rapid anhydride addition, we
might block all of the hydroxyls as adipate esters, preventing
cross-linking. This approach also was unsuccessful; gelation of
the reaction mixture occurred even before adipic anhydride
addition was complete.
If ester cross-linking by condensation of carboxyls on one
cellulose chain with hydroxyls on another cellulose chain was the
problem, then bases like triethylamine and pyridine might be
useful to convert the initial product cellulose adipate carboxylic
acid into a carboxylate anion, making condensation with a
cellulosic hydroxyl to form an ester linkage far less favorable.
However, the use of organic bases was not pursued, because of
our evidence that reaction with poly(adipic anhydride) was in
fact the issue and because of reports40 that triethylamine can
initiate ring-opening polymerization of adipic anhydride to
polyadipic anhydride at room temperature.
The formation of these products was confirmed by proton
NMR studies. Figure 3 shows the 1H NMR spectrum of benzyl
cellulose acetate adipate propionate. One of the adipate peaks
(COCH2CH2CH2CH2CO) appears as a broad singlet at δ 1.66.
The other adipate peak (COCH2CH2CH2CH2CO) merges with
the COCH3 and COCH2CH3 peaks from CAP at δ 2.16À2.35.
The benzyl protons appear at 5.1 (CH2) and 7.33 (Ph). Figure 4
1
shows the H NMR spectrum of the hydrogenation product,
cellulose acetate adipate propionate. The success of the hydro-
genation is confirmed by the disappearance of the benzyl
aromatic signals, and the spectrum clearly indicates that cellulose
acetate adipate propionate has been formed. The 13C NMR
spectrum (Figure 5) also clearly indicates the nearly complete
absence of benzyl signals (aromatic signals 110À130 ppm) and
the presence of the adipate groups. Therefore, the use of the
monofunctional reagent has indeed resulted in smooth prepara-
tion of an un-cross-linked cellulose adipate product, as predicted.
Table 2 shows the results of several attempts to apply this
chemistry to other cellulose derivatives, expanding the range of
adipate derivatives available. MEK was used as a solvent for
reaction of monobenzyl adipoyl chloride with cellulose acetate
propionate (CAP-504-0.2; entry 1, Table 2) and cellulose acetate
butyrate (CAB-553-0.4; entry2, Table 2). The cellulose acetate we
1111
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115

Biomacromolecules
ARTICLE
Figure 3. 1H NMR spectrum of benzyl cellulose acetate adipate propionate in CDCl3.
used (CA-320S) was not soluble in MEK, so reaction of CA was
carried out in the polar aprotic solvent DMI. In each case, rather
low DS commercial cellulose esters were used as starting materials
toprovide a higherDS of availableOHgroups for reaction with the
adipate derivative. Triethylamine was employed as catalyst and
acid scavenger. The product benzyl adipate derivatives of CAP,
CAB, and CA wereeach soluble in THF. Hydrogenation ofeachof
these three products at room temperature using a hydrogen
balloon proceeded smoothly, resulting in benzyl deprotection to
afford the respective cellulose adipate derivatives (2, Scheme 6).
However, hydrogenation of benzyl cellulose adipate (DS
(adipate) = 2.00, entry 5, Table 2) could not be carried out
because of limited solubility of the product in common organic
solvents. Synthesis of cellulose adipate alkanoates from cellulose in
solution in solvents like DMAc/LiCl and ionic liquids should
overcome this product solubility problem. Our results on the
synthesis of cellulose adipates in ionic liquids will be discussed in a
forthcoming paper.
The degrees of substitution (DS) of the products were assigned
from the proton NMR spectra using the ratios of appropriate acyl
proton integrations to the backbone proton integration. For
example, DS (adipate) 0.33 was achieved by reaction between
CAP-504-0.2 and 1 equiv of monobenzyl adipoyl chloride (entry 1,
Table 2). The level of carboxyalkanoyl substituent achieved is
promising for the use of this cellulose adipate derivative for oral
dosage forms in drug delivery or for water-borne dispersions.
For comparison,11 carboxymethylcellulose acetate butyrate (Eastman
CMCAB-641-0.2, DS of carboxymethyl = 0.33) can be readily
dispersed in aqueous media and swells at near-neutral pH.
CMCAB with that DS carboxyl-containing substituent was also
found to provide an excellent matrix for pH-controlled drug
release, zero-order drug release, and solubility enhancement.
Therefore, we anticipate that cellulose adipate alkanoates may
be valuable biopolymers for pharmaceutical formulations. Higher
DS (adipate) can be achieved, although with some reduction in
efficiency; using 3 equiv of monobenzyl adipoyl chloride (entry 2,
Table 2) afforded CAP adipate of DS (adipate) 0.5. Reaction of
cellulose acetate butyrate with 1 equiv of monobenzyl adipoyl
chloride afforded the anticipated cellulose acetate adipate butyrate
with adipate DS 0.25. Reaction of monobenzyl adipoyl chloride
with CA-320S, which has less bulky substituents and a higher
DS(OH) than the other starting esters, afforded the expected
product with a higher adipate DS of 0.7. The benzyl ester products
and their characteristics are summarized in Table 2; deprotected
cellulose ester adipate properties are summarized in Table 3.
For amorphous dispersion formulations of drugs in polymer
matrices, high polymer glass-transition temperature (Tg) is an
advantage.43 If the polymer remains in the glassy state, even at
high ambient temperatures and at high humidity, then the
mobility of the drug molecules is restricted and their crystal-
lization is suppressed. Therefore, it was of interest to investigate
the thermal properties of our cellulose adipate esters. Table 3
1112
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115

Biomacromolecules
ARTICLE
Figure 4. 1H NMR spectrum of cellulose acetate adipate propionate in CDCl3.
Figure 5. 13C NMR spectrum of cellulose acetate adipate propionate in DMSO-d6.
1113
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115

Biomacromolecules
ARTICLE
Table 3. Properties of Starting Materials and Deprotected Cellulose Ester Adipates
starting material data
product data
Mn/103 (SEC)
entry
starting material
Tg 2nd heat(°C)
Tm (°C)
Mn/103 (SEC)a
product
Tg 2nd heat (°C)
solubility
1
2
3
CAP-504-0.2
CAB-553-0.4
CA-320S
158
100
180
188
15.0
20.0
38.0
CAP adipate
CAB adipate
CA adipate
À14, 125
À18, 94
À18, 131
10.8
CHCl3, NMP, DMSO
CHCl3, NMP, DMSO
NMP, DMSO
155
61.0
b
230À250
a Reported by supplier, versus polystyrene standards. b Mn could not be measured by SEC.
of cellulose in DMAC/LiCl solution or cellulose esters in MEK
solution with a monofunctional adipate reagent, monobenzyl
adipoyl chloride, followed by removal of the benzyl group by
mild hydrogenation. We have observed no limitation on the type
of alkanoate group that may be present. In some cases, we have
achieved high DS of adipate, so there is no apparent limitation on
achievable adipate DS. Suitable starting materials include cellu-
lose and esters of cellulose; ethers and other cellulose derivatives
are also likely to be suitable. This process circumvents the cross-
linking problems seen when using adipic anhydride and leads to
soluble cellulose (ester) adipate derivatives. We anticipate that
these new cellulose derivatives will be useful in aqueous coatings
systems and for controlled release of drugs and other active
molecules. Studies of the utility of these cellulose adipate esters in
drug delivery are under way.
’ ASSOCIATED CONTENT
Figure 6. DSC curves of cellulose adipate derivatives at heating rate
10 °C/min (2nd heat).
Supporting Information. Complete infrared, 13C NMR,
S
b
1
and H NMR spectra for the new cellulose derivatives synthe-
summarizes the results of differential scanning calorimetry
(DSC) studies of the respective adipates, and Figure 6 displays
the DSC traces. All cellulose ester adipates prepared had glass-
transition temperatures in excess of 90 °C, and the CA and CAP
adipates had Tg well in excess of 120 °C, indicating promise for
amorphous dispersion formulations. We also observed low-
temperature transitions for these cellulose ester adipates, well
below ambient temperature, which were not evident in the
starting materials. We speculate that these transitions may be
due to cooperative motions of the relatively long adipate side
chains, similar to those observed by Glasser and coworkers for
cellulose long-chain esters.44
sized in this work, other than those included in the body of the
manuscript, the DSC spectrum of cellulose acetate adipate
propionate, and 1H NMR spectra of the monofunctional adipate
reagents prepared for this study. This material is available free of
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: kjedgar@vt.edu. Tel: 540.231.0674.
’ ACKNOWLEDGMENT
Molecular weight characterization of these cellulose adipate
materials was challenging. We first explored size-exclusion chro-
matography (SEC) in an NMP/LiCl solvent system but did not
observe polymer signals in the column eluant. Some samples
showed aggregation by laser light scattering studies in DMF
solvent, whereas other samples did not. Apparently these cellu-
lose adipates are somewhat prone to aggregation and have high
affinity for the polystyrene beads of the SEC column. Finally, we
found that SEC in DMF containing 0.05 M LiBr was successful
for the CAP and CAB adipates; the less-soluble CA was still
problematic in this system. In general, the Mn results (Table 3)
obtained indicate no more than moderate chain degradation
during the adipate ester preparation, consistent with the rela-
tively mild conditions used.
We thank the Eastman Chemical Company and Rayonier for
their kind donations of the cellulose ester starting materials and
high-purity hardwood dissolving pulp used in this work, respec-
tively. We thank the National Science Foundation and the U.S.
Department of Agriculture for funding this work and the
Macromolecules and Interfaces Institute and the Institute for
Critical Technologies and Applied Science at Virginia Tech for
their support. We also thank Tianyu Wu and Matt Green of
Virginia Tech for performing SEC analyses.
’ REFERENCES
(1) Buchanan, C. M.; Edgar, K. J.; Hyatt, J. A.; Wilson, A. K.
Macromolecules 1991, 24, 3050–3059.
(2) Kamitakahara, H.; Nakatsubo, F.; Klemm, D. Cellulose 2006,
13, 375–392.
’ CONCLUSIONS
(3) Kamitakahara, H.; Koschella, A.; Mikawa, Y.; Nakatsubo, F.;
Heinze, T.; Klemm, D. Macromol. Biosci. 2008, 8, 690–700.
(4) Heinze, T.; Koschella, A. Macromol. Symp. 2005, 223, 13–39.
We have developed a general process for the synthesis of
cellulose adipates, with or without other ester groups, by reaction
1114
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115

Biomacromolecules
ARTICLE
(5) Arisz, P.; Thai, H. T. T.; Boon, J. J.; Salomons, W. G. Cellulose
1996, 3, 45–61.
(6) Rekhi, G. S.; Jambhekar, S. S. Drug Dev. Ind. Pharm. 1995,
21, 61–77.
(39) Mathiowitz, E.; Jacob, J. S.; Jong, Y. S.; Carino, G. P.; Chickering,
D. E.; Chaturvedi, P.; Santos, C. A.; Vijayaraghavan, K.; Montgomery, S.;
Bassett, M.; Morrell, C. Nature 1997, 386, 410–4.
(40) Albertsson, A. C.; Carlfors, J.; Sturesson, C. J. Appl. Polym. Sci.
(7) Koschella, A.; Fenn, D.; Heinze, T. Polym. Bull. 2006, 57, 33–41.
(8) Junior, J. A.; Kawano, Y.; Petri, D. F. S. Mater. Sci. Eng., C 2009,
29, 420–425.
1996, 62, 695–705.
(41) Nakatsubo, F.; Kamitakahara, H.; Hori, M. J. Am. Chem. Soc.
1996, 118, 1677–1681.
(9) Edgar, K. J. Cellulose 2007, 14, 49–64.
(42) Kamitakahara, H.; Nakatsubo, F.; Klemm, D. Cellulose 2007,
14, 513–528.
(43) Weuts, I.; Van Dycke, F.; Voorspoels, J.; De Cort, S.; Stokbroekx,
S.; Leemans, R.; Brewster, M. E.; Xu, D.; Segmuller, B.; Turner, Y. T.;
Roberts, C. J.; Davies, M. C.; Qi, S.; Craig, D. Q.; Reading, M. J. Pharm. Sci.
2011, 100, 260–74.
(10) Friesen, D. T.; Shanker, R.; Crew, M.; Smithey, D. T.; Curatolo,
W. J.; Nightingale, J. A. Mol. Pharm. 2008, 5, 1003–19.
(11) Posey-Dowty, J. D.; Watterson, T. L.; Wilson, A. K.; Edgar,
K. J.; Shelton, M. C.; Lingerfelt, L. R. Cellulose 2007, 14, 73–83.
(12) Shelton, M. C.; Posey-Dowty, J. D.; Lingerfelt, L. R.; Kirk, S. K.;
Klein, S.; Edgar, K. J., Enhanced Dissolution of Poorly Soluble Drugs
from Solid Dispersions in Carboxymethylcellulose Acetate Butyrate
Matrices. In Polysaccharide Materials: Performance by Design; Edgar,
K. J., Heinze, T., Liebert, T., Eds.; American Chemical Society:
Washington, D.C., 2009.
(44) Sealey, J. E.;Samaranayake, G.;Todd, J. G.;Glasser, W. G. J. Polym.
Sci., Part B: Polym. Phys. 1996, 34, 1613–1620.
(13) Alonzo, D. E.; Zhang, G. G.; Zhou, D.; Gao, Y.; Taylor, L. S.
Pharm. Res. 2010, 27, 608–18.
(14) Konno, H.; Taylor, L. S. J. Pharm. Sci. 2006, 95, 2692–705.
(15) Konno, H.; Taylor, L. S. Pharm. Res. 2008, 25, 969–78.
(16) Konno, H.; Handa, T.; Alonzo, D. E.; Taylor, L. S. Eur. J. Pharm.
Biopharm. 2008, 70, 493–9.
(17) Leuner, C.; Dressman, J. Eur. J. Pharm. Biopharm. 2000,
50, 47–60.
(18) Klein, C. E.; Chiu, Y. L.; Awni, W.; Zhu, T.; Heuser, R. S.; Doan,
T.; Breitenbach, J.; Morris, J. B.; Brun, S. C.; Hanna, G. J. JAIDS, J.
Acquired Immune Defic. Syndr. 2007, 44, 401–10.
(19) Allen, J. M.; Wilson, A. K.; Lucas, P. L.; Curtis, L. G. Carboxyalkyl
Cellulose Esters. Patent 5,668,273, September 16, 1997.
(20) Allen, J. M.; Wilson, A. K.; Lucas, P. L.; Curtis, L. G. Process for
Preparing Carboxyalkyl Cellulose Esters. Patent 5,792,856, August 11,
1998.
(21) Posey-Dowty, J. D.; Wilson, A. K.; Curtis, L. G.; Swan, P. M.;
Seo, K. S. Carboxyalkyl Cellulose Esters for Use in Aqueous Pigment
Dispersions. Patent 5,994,530, November 30, 1999.
(22) Cook, P. M.; Smith, J. V. Cellulose Acetate Phthalate Enteric
Coating Compositions. Patent 5,888,550, March 30, 1999.
(23) Cook, P. M.; Lambert, J. L. pH-Sensitive Modified Cellulose
Ester. Patent 5,925,181, July 20, 1999.
(24) Edgar, K. J. Polym. Paint Colour J. 1993, 183, 564–571.
(25) Seidel, C.; Kulicke, W.-M.; Hess, C.; Hartmann, B.; Lechner,
M. D.; Lazik, W. Starch/Staerke 2001, 53, 305–310.
(26) Coma, V.; Sebti, I.; Pardon, P.; Pichavant, F. H.; Deschamps, A.
Carbohydr. Polym. 2003, 51, 265–271.
(27) Mitchell, G. R.; Guo, W.; Davis, F. J. Polymer 1992, 33, 68–74.
(28) Thomas, R. Textilveredlung 1970, 5, 361–368.
(29) Morello, A. P., 3rd; Forbes, N.; Mathiowitz, E. J. Microencapsul.
2007, 24, 40–56.
(30) Carnahan, M. A.; Grinstaff, M. W. Macromolecules 2006,
39, 609–616.
(31) Albertsson, A.-C.; Lundmark, S. J. Macromol. Sci., Chem. 1988,
A25, 247–258.
(32) English, A. R.; Girard, D.; Jasys, V. J.; Martingano, R. J.; Kellogg,
M. S. J. Med. Chem. 1990, 33, 344–347.
(33) Abell, A. D.; Morris, K. B.; Litten, J. C. J. Org. Chem. 1990,
55, 5217–5221.
(34) Edgar, K. J.; Arnold, K. M.; Blount, W. W.; Lawniczak, J. E.;
Lowman, D. W. Macromolecules 1995, 28, 4122–4128.
(35) Dawsey, T. R.; McCormick, C. L. Polym. Rev. 1990,
30, 405–440.
(36) El Seoud, O. A.; Koschella, A.; Fidale, L. C.; Dorn, S.; Heinze, T.
Biomacromolecules 2007, 8, 2629–47.
(37) McCormick, C. L.; Dawsey, T. R. Macromolecules 1990,
23, 3606–3610.
(38) Ron, E.; Turek, T.; Mathiowitz, E.; Chasin, M.; Hageman, M.;
Langer, R. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 4176–80.
1115
dx.doi.org/10.1021/bm101448f |Biomacromolecules 2011, 12, 1106–1115