Stereoregularity of poly(lactic acid) and their model compounds as studied by NMR and quantum chemical calculations

Article abstract of DOI:10.1021/ma2018777

In order to understand the origin of the tacticity splitting in the NMR spectra of poly(lactic acid), monomer and dimer model compounds were synthesized and their ¹H and ¹³C NMR chemical shifts were observed. Two stable conformations were obtained from Ramachandran map calculated as a function of the internal rotation angles for the monomer model using Gaussian 09 calculations. Four preferred conformations were selected and optimized for each dimer model. The conformations of neighboring residues were energetically interdependent. The ¹H and ¹³C chemical shifts for dimer model compounds were calculated by averaging the occurrence probabilities obtained from the optimized conformational energies and the calculated chemical shift of each conformation. It was confirmed that the solvent effect on the tacticity-dependent relative chemical shifts was small from the NMR experiments of the model compound observed in different solvents, dimethyl sulfoxide, chloroform, and chloroform/carbon tetrachloride (20/80 v/v) mixture. Good agreement between observed and calculated chemical shifts was obtained for the relative chemical shifts of isotactic and syndiotactic ¹H and ¹³C NMR peaks of the dimer model compounds. The observed tacticity splitting of poly(lactic acid) at the diad level was rationalized on the basis of these chemical shift calculations.

Full text of DOI:10.1021/ma2018777

ARTICLE
pubs.acs.org/Macromolecules
Stereoregularity of Poly(lactic acid) and their Model Compounds as
studied by NMR and Quantum Chemical Calculations
†
,‡
‡
‡
§
†
,†
Koto Suganuma, Ken Horiuchi, Hironori Matsuda, H. N. Cheng, Akihiro Aoki, and Tetsuo Asakura*
†
Department of Biotechnology, Tokyo University of Agriculture and Technology, Koganei, Tokyo, 184-8588, Japan
‡
Material Analysis Research Laboratories, Teijin Ltd., Hino, Tokyo, 191-8512, Japan
§
Agriculture Research Service, Southern Regional Research Center, United States Department of Agriculture, 1100 Robert E. Lee Blvd.,
New Orleans, Louisiana 70124, United States
ABSTRACT: In order to understand the origin of the tacticity splitting in
the NMR spectra of poly(lactic acid), monomer and dimer model com-
1
13
pounds were synthesized and their H and C NMR chemical shifts were
observed. Two stable conformations were obtained from Ramachandran
map calculated as a function of the internal rotation angles for the monomer
model using Gaussian 09 calculations. Four preferred conformations were
selected and optimized for each dimer model. The conformations of
1
13
neighboring residues were energetically interdependent. The H and
C
chemical shifts for dimer model compounds were calculated by averaging the occurrence probabilities obtained from the optimized
conformational energies and the calculated chemical shift of each conformation. It was confirmed that the solvent effect on the
tacticity-dependent relative chemical shifts was small from the NMR experiments of the model compound observed in different
solvents, dimethyl sulfoxide, chloroform, and chloroform/carbon tetrachloride (20/80 v/v) mixture. Good agreement between
1
13
observed and calculated chemical shifts was obtained for the relative chemical shifts of isotactic and syndiotactic H and C NMR
peaks of the dimer model compounds. The observed tacticity splitting of poly(lactic acid) at the diad level was rationalized on the
basis of these chemical shift calculations.
’
INTRODUCTION
and “s” syndiotactic pairwise relationships (-LD- and -DL-). In the
NMR spectra, the diads -LL- and -DD- are indistinguishable and
have the same chemical shifts, but as do the diads -LD- and -DL-.
Poly(lactic acid) (PLA) can exist in amorphous and semicrys-
talline forms in the solid state. In blends of poly(L-lactic acid)
PLLA) and poly(D-lactic acid) (PDLA), the two molecular
12
Bero et al. assigned all the CH carbon peaks at the tetrad level
and part of the carbonyl peaks at the hexad level, using PLA
samples prepared under various reaction conditions and different
(
species can form a 1/1 stereocomplex, and the crystalline
structure of the stereocomplex is quite different from that of
PLLA or PDLA.
temperature higher than that of the homopolymer by 50 °C.
Such a high melting temperature was reported first from a sample
isolated from a solution and later from a melt mixture. Melt
blending of PLLA and PDLA is likely accompanied by the
formation of single polymer crystals together with stereocomplex
1
3
1
1
À5
initiators. Kricheldorf et al. reported the H assignments of the
13
The stereocomplex PLA has a melting
CH group of PLA on the basis of the previous C assignments
with a number of PLA samples prepared under various condi-
tions and using numerous initiators/catalysts. Kasperczyk
1
4
2,3
added more assignments to the previous partial hexad assign-
ments of the carbonyl group of PLA. Partial hexad level assign-
1
13
6
À8
9
ments of the H and C peaks of the CH group were reported by
15
formation, resulting in a mixed crystalline state.
Yui et al.
Thakur et al. with PLA samples prepared in vials by melt
polymerization of various combinations of L-lactide, D-lactide,
discovered that a diblock copolymer of PLLAÀPDLA could
form the stereocomplex easily because of the neighboring effect
of the two blocks in the copolymer. On the basis of this result,
Fukushima et al. reported that multiblock copolymers of PLLA
and PDLA formed the stereocomplex more easily without first
forming the single-polymer crystals. However, PLA with short
L- and D-block sequences caused lower melting temperature.
Thus, it is important to have an improved knowledge of stereo-
regularity and sequence distribution in PLA samples.
17
and meso-lactide. Chisholm et al. reported further investigation
1
13
1
0
of the CH H and C assignments with two-dimensional (2D)
NMR such as HETCOR and claimed that previous tetrad level
1
1
3
6
assignments for the CH carbon reported by Kricheldorf et al.
11
were partly in error; This claim was disputed by Kasperczyk
18
18
and Zell et al. In addition, Zell et al. also reported NMR data
from HMQC and HMBC experiments and partly C labeled
PLA samples and made improved assignments of (sii) and (iis)
1
3
NMR is generally regarded as the best method to study
polymer stereoregularity. Many NMR studies have been re-
12À18
ported of stereosequence distribution in PLA.
It is common
Received: August 17, 2011
in the PLA literature to designate the assignments as various
Revised:
October 23, 2011
combinations of “i” isotactic pairwise relationships (-LL- and -DD-)
Published: November 03, 2011
r 2011 American Chemical Society
9247
dx.doi.org/10.1021/ma2018777
|Macromolecules 2011, 44, 9247–9253

Macromolecules
ARTICLE
tetrad peaks of the CH proton. Using these several techniques,
1
13
they reported unambiguous determination of the H and
tacticity assignment of PLA.
C
18
The origin of the tacticity splitting of chemical shifts has been
studied for many polymers through semiempirical chemical shift
1
3
calculation by considering the C γ-gauche shielding effect or
Figure 1. Monomer model compound (1) of poly(lactic acid).
1
19À22
H ring current shielding effect.
Since the time-averaged
occurrence of several conformations of polymer chains in solu-
tion must be considered in the chemical shift calculation, it is
useful to carry out the conformational energy calculation of
model compounds or fragments of the polymer chain. There
have been two approaches used, viz., semiempirical energy
22
calculation and quantum chemical calculation. Brant et al.
Figure 2. Dimer model compound (2) of poly(lactic acid).
calculated the conformational energy of the PLLA chain using
the former method and prepared a rotational isomeric state
model of PLLA from the conformational energy maps to
reproduce flexible PLA chain and to explain the solution property
of the polymer. Meaurio et al. analyzed the carbonyl stretching
region of IR spectrum of PLA and discussed the conformational
population via simulation of the vibrational behavior of the
polymer chain.
Water (2 mL) and tetrahydrofuran (THF) (5 mL) were added to the
mixture. The mixture was thoroughly mixed and transferred to a
separatory funnel. The two layers (aqueous and toluene) were allowed
to separate, and the aqueous layer was discarded. The toluene layer was
washed with aqueous sodium hydrogen carbonate solution and passed
through a bed of sodium sulfate to remove any residual water. The
toluene was evaporated using a rotary evaporator at 40 °C. The product
was purified by size exclusion chromatography.
23
The quantum chemical calculations are potentially more
accurate and are becoming less time-consuming to do because
24À27
of increasing computer speed and memory size.
Two
Synthesis of Dimer Model Compound (2) of Poly(lactic acid).
The dimer model compound (2) shown in Figure 2 was synthesized as
follows.
conformational analyses of PLA were reported using quantum
28
chemical calculations. The conformational energies of PLA
oligomers were calculated using a quantum chemical method by
2
Thionyl chloride (SOCl ) (238 mg, 2 mmol) and DMF (15.5 μL)
2
9
McAliley et al., in vacuo or in the electronic environment within
were added to a solution of (-)-O-acetyl-L-lactic acid (132 mg, 1.0 mmol)
in toluene (660.6 μL). The mixture was stirred at 80 °C for 1 h, and the
mixture was evaporated using a rotary evaporator at 40 °C. The solution
of methyl S-(À)-lactate (83 mg, 1.0 mmol) or methyl R-(À)-lactate (83
mg, 1.0 mmol) in toluene (660.6 μL) and pyridine (161.1 μL) was added
to the mixture. The mixture was stirred at room temperature for 2 h.
Aqueous sodium hydrogen carbonate solution (2 mL) and tetrahydro-
furan (THF) (5 mL) were added to the mixture. The mixture was
thoroughly mixed and transferred to a separatory funnel. The two layers
26
the condensed phase. Wu et al. determined the equilibrium
geometries, IR, total energies and NMR chemical shifts of several
lactides, including L-lactide, D-lactide, and three meso-lactides. It
may be noted that lactides have a ring structure and are different
from the structure of polymeric PLA.
In this work, in order to understand the origin of the tacticity
splitting in the NMR spectra of PLA, we synthesized the
monomer model compound and isotactic and syndiotactic dimer
model compounds of PLA and compared the observed and the
(aqueous and toluene) were allowed to separate, and the aqueous layer
1
13
1
13
was discarded. The toluene layer was washed with aqueous sodium
hydrogen carbonate solution and passed though a bed of sodium sulfate
to remove any residual water. The toluene was evaporated using a rotary
evaporator at 40 °C. The product was purified by size exclusion
chromatography.
Preparation of Poly(lactic acid). Copolymer of poly(L-lactic
acid) (PLLA) and poly(D-lactic acid) (PDLA), at 50/50 mol %, was
prepared by transesterification between PLLA and PDLA. Heat treat-
ment of the blend of PLLA and PDLA with sodium benzoate (1 wt %)
was performed on a TG-DTA Instruments (Rigaku Thermo Plus TG
calculated H and C NMR chemical shifts of all H and
C
nuclei in the models. For comparison, the solvent effect on the
tacticity-dependent relative chemical shifts was examined for the
dimer model compounds. For the chemical shift calculation, the
conformational energy calculation was performed for each iso-
tactic and syndiotactic dimer model compound on the basis of
the combination of two energetically stable conformations
obtained for the monomer model compound by quantum
1
13
chemical calculations. The H and C chemical shifts of all the
1
13
H and C nuclei were also calculated for each conformation of
the model compounds by the quantum chemical method. The
8
120) under dry nitrogen atmosphere. A sample was heated from room
À1
temperature to 220 °C with a heating rate of 20 K min , heated to
1
13
chemical shifts were then obtained for all the H and C nuclei in
isotactic and syndiotactic dimer model compounds. On the basis
of the calculated and observed dimer data, an attempt was made
to rationalize the tacticity splitting of the CH region of the PLA
spectra at the diad level.
À1
2
50 °C with a rate of 10 K min , maintained at 250 °C for 60 min, and
quenched into liquid nitrogen. The product was washed with methanol.
1
13
NMR Measurement. The H and C NMR spectra were obtained
on a JEOL α-600 spectrometer operating at 600 and 150 MHz,
respectively, at room temperature. Deuterated chloroform (CDCl
deuterated chloroform/carbon tetrachloride (20/80 in vol %) mixture
CDCl /CCl ) and deuterared dimethyl sulfoxide (DMSO-d ) were
3
),
(
’
EXPERIMENTAL SECTION
3
4
6
used as the solvents and the sample concentration was 10% (w/v).
Tetramethylsilane (TMS) was used as an internal standard chemical
Synthesis of Monomer Model Compound (1) of Poly(lactic
acid). The monomer model compound (1) shown in Figure 1 was
synthesized as follows.
Pyridine (523.6 μL) and acetyl chloride (471 mg, 6.0 mmol) were
added to a solution of methyl S-(À)-lactate (521 mg, 5.0 mmol) in
toluene (2.6 mL). The mixture was stirred at room temperature for 2 h.
1
shift reference. The H NMR spectra were obtained with a digital
resolution of 0.36 Hz/point with 32 K data points together with 45° flip
1
3
angle and 4s pulse delay. The C NMR spectra were obtained with a
digital resolution of 1.24 Hz/point with 32 K data points together with
45° flip angle and 2 s pulse delay.
9
248
dx.doi.org/10.1021/ma2018777 |Macromolecules 2011, 44, 9247–9253

Macromolecules
ARTICLE
Figure 4. Ball-and-stick models of two energetically preferred confor-
mations of monomer model compound (1). (a) The most stable
conformation, (Φ,Ψ) = (À66°, 155°). (b) The second stable con-
formation, (Φ,Ψ) = (À82°, 7°).
1
13
Table 1. H (Upper) and C (Lower) Chemical Shifts (ppm)
from TMS of the Monomer Model Compound (1) together
with the Energy Difference (ΔE) and Probability (%)
Figure 3. Ramachandran map of the conformational energies as a
function of the internal rotation angles, Φ and Ψ for molecule 1.
Conformational Energy Calculation. The conformational en-
ergy calculations of monomer model compound (1) were carried out
with Gaussian 09 software as a function of the internal rotation angle.
30À32
The basis set was TZVP.
The conformational energy calculations
were also performed for isotactic and syndiotactic dimer model com-
pounds for the preferred conformations selected from the Ramachan-
dran map of the monomer model compound.
calcd
conformation a conformation b average
obsd
Chemical Shift Calculation. The magnetic shielding tensor was
3
3À37
calculated quantum-chemically according to the GIAO method
for
H and C nuclei of the two preferred conformations of the monomer
model compound (1). The isotropic chemical shifts were obtained for
ΔE (kJ/mol)
0.00
74.3
2.00
2.66
25.7
1.96
1
13
probability (%)
1
H
1
4
7
8
1
2
4
5
7
8
2.00
4.87
2.12
5.08
1
13
comparison with the observed chemical shifts. The H and C chemical
shifts of TMS were calculated with the same method. As for the H and
C chemical shift calculations for dimer model compounds, a two-step
4.88
4.84
1
3.61
3.56
3.60
3.74
1
3
1.37
1.39
1.38
1.48
procedure was used. In each model compound, the relative occurrence
probabilities of the preferred conformations were first calculated by
Boltzmann distribution on the basis of the difference in the conforma-
tional energy. Since the chemical shifts were calculated for each
conformation quantum-chemically, the chemical shift of each config-
uration could be obtained by taking into account both the relative
occurrence probability and the chemical shifts of each conformation.
1
3
C
20.82
177.50
72.59
180.05
53.62
17.95
20.93
177.48
74.35
179.78
53.78
18.53
20.84
20.31
177.50 170.00
72.67 68.25
179.98 171.00
53.66
18.01
51.97
20.31
although the deviation in Ψ value was slightly larger. There were
additional minima at (À160°, 160°) and (À160°, À48°) in their
model which was not predicted in our calculations. However, the
fractions of the conformations in the map reported by Brant et al.
were 55% for (Φ,Ψ) = (À73°, 160°), 37% for (Φ,Ψ) = (À73°,
À48°), and sum of other two conformations was 8% at room
temperature. In our case, the fraction of the conformations is 74%
for (Φ,Ψ) = (À66°, 155°) and 26% for (Φ,Ψ) = (À82°, 7°)
at room temperature. Thus, the difference in the maps between
the results by Brant et al. and this work is relatively small although
the method of calculations seems quite different. The two
energetically stable states at (Φ,Ψ) values of (À66°, 155°) and
(À82°, 7°) are shown in Figure 4 as ball-and-stick models.
Although the position of oxygen atom changes almost in the
opposite direction in the molecules due to the difference in the
internal rotation angle of Ψ, the basic arrangements of the inner
backbone atoms are about the same.
’
RESULTS AND DISCUSSION
1). Ramachandran Map of Monomer Model Compound
(
(
1) as a Function of the Internal Rotation Angles, Φ and Ψ
The conformational energies of monomer model compound, 1,
were calculated with Gaussian 09 as a function of the internal
rotation angles, Φ and Ψ, and plotted as shown in Figure 3.
According to our calculations, there are two energetically
stable states at (Φ,Ψ) values of (À66°, 155°) and (À82°, 7°).
29
Earlier, McAliley et al. calculated the conformational energies
of PLA trimer model as a function of the internal rotation angles,
Φ and Ψ, by a quantum chemical method (electron density
functional theory at the B3LYP/6-31G** level) both in vacuo and
with a self-consistent reaction field method. Although the model
and the details of the quantum chemical methods are different
between McAliley et al. and this work, similar Ramachandran
maps are obtained. In contrast, the conformational energy map
22
1
13
reported by Brant et al. was slightly different. In their map,
there were four energetically stable states. The most stable state
was the (Φ,Ψ) value of (À73°, 160°), which was similar to the
value obtained in this work (Φ,Ψ) = (À66°, 155°). The next
energetically stable state was (Φ,Ψ) = (À73°, À48°), which was
similar to the value obtained in this work (Φ,Ψ) = (À82°, 7°)
(2). H and C Chemical Shift Calculations of Monomer
1
13
Model Compound (1). The H and C chemical shifts were
1
13
calculated with the quantum chemical method for all H and C
nuclei of the monomer model compound (1) with only two
conformations, (a) (Φ,Ψ) = (À66°, 155°), and (b) (À82°, 7°) .
1
13
The calculated H and C chemical shifts (ppm) are listed in
9
249
dx.doi.org/10.1021/ma2018777 |Macromolecules 2011, 44, 9247–9253

Macromolecules
ARTICLE
Table 1 together with the energy difference (ΔE, in kJ/mol), and
probabilities were 0.55, 0.19, 0.19, and 0.07 for L1L1, L1L2,
L2L1, and L2L2, respectively, but 0.77, 0.18, 0.03, and 0.02,
respectively, for the final dimer model compound conforma-
tional probabilities. Similar (and larger) differences are seen for
the syndiotactic dimer model compound. The products of the
monomer model compound conformational probabilities were
0.55, 0.19, 0.19, and 0.07 for L1D1, L1D2, L2D1, and L2D2,
respectively, but 0.13, 0.31, 0.45, and 0.11, respectively, for the
final dimer model compound conformational probabilities. The
larger differences in the syndiotactic case are probably due to the
energetics of the different asymmetric carbon atoms in the syndio-
tactic dimer model compounds. In any case, the conformations of
neighboring PLA residues are energetically interdependent.
probability (%).
1
13
The observed H and C NMR chemical shifts of the
monomer model (1) (also listed in Table 1) were obtained in
deuterated chloroform. Spectral assignments were carried out
with the usual NMR procedures. The calculated chemical shifts
are plotted against the observed ones in Figure 5.
1
High correlation coefficients are obtained, viz., 0.9996 ( H
13
NMR) and 0.9999 ( C NMR).
(3). Conformational Energy Calculations of the Dimer
Model (2). The conformational energies of the four conforma-
tions for each isotactic and syndiotactic dimer model compound
(
2) were calculated by the quantum chemical method. The
1
13
combinations of energetically stable states of two preferred
conformations selected from monomer model (1) were the
initial four conformations of the dimer model (2). The D-isomer
of syndiotactic dimer model was derived by flipping the sign of
the Z coordinate of the XÀYÀZ coordinate. The initial (Φ,Ψ)
and the final (Φ,Ψ) values after optimization by quantum
chemical calculations are summarized in Table 2. The conforma-
tional energies of the final conformations are also listed in
Table 3.
(4). H and C Chemical Shift Calculations of Dimer Model
(2) and Comparison between the Calculated and Observed
1
13
Chemical Shifts. The H and C chemical shifts of each
1
conformation of the dimer model (2) were calculated. The H
13
1
13
and C chemical shifts (ppm) of all H and C nuclei except for
two terminal CH groups are listed in Table 3 together with the
3
energy difference (ΔE, in kJ/mol) and the conformational
probabilities (%). The observed chemical shifts were obtained
1
13
through H and C NMR analysis of the dimer model in
It is noted that there are significant differences between the
products of the monomer model compound conformational
probabilities and the final dimer model compound conforma-
tional probabilities, even though the conformation of each dimer
residue is similar to those derived for the monomer model
compounds. Specifically, for isotactic dimer model compound,
the products of the monomer model compound conformational
DMSO-d , CDCl /CCl (20/80 in vol %) mixture and CDCl .
6
3
4
3
The peak assignments were achieved readily with usual NMR
procedures as well as comparison with the monomer model
1
13
compound. The observed H and C chemical shifts are shown
as stick spectra (Figures 6 and 7, respectively). In the literature,
the NMR solvent of PLA was usually CDCl , but DMSO-d was
3
6
used occasionally although the solvent polarity is quite different.
CCl is an inert solvent but PLA is insoluble in CCl . However,
4
4
the dimer model compound is soluble in CDCl /CCl (20/80 in
3
4
vol %) mixture, where CDCl is used also for NMR field locking.
3
Although the characteristics of three solvent systems are quite
different, the relative tacticity splitting of the dimer model
compounds shows no significant difference among the solvents.
This means that the relative tacticity splitting is due to inherent
difference between two compounds, and we can compare the
observed chemical shifts with the calculated chemical shifts in
vacuo.
Figure 5. Comparison of the calculated and observed chemical shifts of
1
13
(a) H and (b) C nuclei of monomer model compound (1) of
Note that the observed chemical shift differences are very
small between isotactic and syndiotactic dimer models for
poly(lactic acid).
Table 2. Initial Internal Rotation Angles (Φ , Ψ , Φ , Ψ ) and Final Ones (Φ , Ψ , Φ , Ψ ) of the Dimer Model Compound (2)
1
1
2
2
1
1
2
2
after Optimization by Gaussian 09 Software
isotactic type for conformation
syndiotactic type for conformation
L1L1
L1L2
L2L1
L2L2
L1D1
L1D2
L2D1
L2D2
initial (deg)
Φ
1
À66.0
155.0
À66.0
155.0
À69.6
160.3
À73.9
166.8
À66.0
155.0
À82.0
7.0
À82.0
7.0
À82.0
7.0
À66.0
155.0
À66.0
155.0
82.0
À82.0
7.0
À82.0
7.0
Ψ
1
Φ
2
À66.0
155.0
À134.5
50.4
À82.0
7.0
66.0
66.0
82.0
Ψ
2
À155.0
À84.5
À175.3
73.7
À7.0
À76.1
173.2
72.1
À155.0
À70.5
À16.6
74.1
À7.0
À71.9
À13.8
75.6
final (deg)
Φ
1
À71.1
162.8
À74.9
À11.4
À81.9
À10.6
À74.9
À11.0
Ψ
1
Φ
2
À71.5
165.5
Ψ
2
À162.3
11.6
À166.6
11.2
9
250
dx.doi.org/10.1021/ma2018777 |Macromolecules 2011, 44, 9247–9253

Macromolecules
ARTICLE
1
13
Table 3. H (Upper) and C (Lower) Chemical Shifts (ppm) from TMS of the Isotactic and Syndiotactic Dimer Model
a
Compounds (2) together with the Energy Difference (ΔE) and Probability (%) of Each Conformation
isotactic type for conformation
calcd
syndiotactic type for conformation
calcd
L1L1
0.00
L1L2
3.71
L2L1
L2L2
average
obsd
L1D1
3.11
L1D2
L2D1
L2D2
average
obsd
ΔE (kJ/mol)
8.03
9.04
0.96
0.00
3.45
probability (%)
77.3
4.47
17.5
4.51
3.1
5.39
2.1
4.87
13.0
4.51
30.6
4.78
45.1
4.45
11.3
4.51
1
H
4
7
1
1
4
7
1
1
2
5
8
4.48
4.57
5.11
5.17
4.52
4.51
5.16
5.15
4.56
1.58
4.60
1.53
4.46
1.35
4.56
1.44
4.05
1.17
4.54
1.46
4.48
1.55
4.55
1.52
1
2
1.56
1.57
1.50
1.53
1.46
1.48
1.42
1.43
1.46
1.53
1.14
1.45
1.38
1.41
1.39
1.51
1
3
C
66.01
65.37
16.65
16.61
182.87
182.82
183.55
65.50
66.62
16.57
17.38
182.71
182.69
183.67
64.33
65.78
16.42
16.53
181.45
182.53
182.83
65.11
66.96
17.65
17.23
182.11
183.82
183.34
65.96
65.46
16.63
16.63
182.84
182.79
183.56
68.30
69.00
16.68
16.75
170.33
170.24
170.65
59.61
61.12
15.78
15.60
172.31
173.40
173.50
64.76
67.04
16.55
17.13
182.46
182.38
183.41
66.92
65.35
17.14
16.42
182.96
182.79
183.11
66.42
66.71
17.23
17.17
182.82
183.06
183.65
66.22
66.09
17.05
16.58
182.80
182.68
183.28
68.44
69.14
16.77
16.80
170.14
170.08
170.44
1
2
a
The observed chemical shifts in CDCl are listed here; more chemical shift data are also shown in Figures 6 and 7.
3
1
3
Figure 7. Comparison of the observed and calculated C chemical
shifts (in ppm) of dimer model compound (2) of poly(lactic acid)
1
6 3 4
shown as stick spectra. The solvent is (a) DMSO-d , (b) CDCl /CCl
Figure 6. Comparison of the observed and calculated H chemical shifts
in ppm) of dimer model compound (2) of poly(lactic acid) shown as
stick spectra. The solvent is (a) DMSO-d , (b) CDCl /CCl (20/80 in
vol %) mixture, and (c) CDCl . The calculated chemical shifts are shown
(
20/80 in vol %) mixture, and (c) CDCl . The calculated chemical shifts
(
3
are shown as part d. The chemical shifts are shown relative to the
isotactic chemical shift. The blue stick corresponds to isotactic and the
pink syndiotactic.
6
3
4
3
as (d). The chemical shifts are shown relative to the isotactic chemical
shift. The blue stick corresponds to isotactic and the pink syndiotactic.
H-4 of isotactic dimer model, the value for L2L1 is remarkably
large; however, its probability is only 3.1% and therefore
negligible when all conformations are considered. The chemical
shift difference between H-4 and H-7 in L1L1 with the occur-
rence probability 77.3% seems to be the origin of the reversed
peak positions between syndiotactic and isotactic dimer models
(Table 3). In contrast, the observed relative peak positions are
poly(lactic acid) for both nuclei when compared with the
13,14
chemical shift difference in other polymers.
It is instructive
to compare the relative peak position for each nucleus between
the isotactic and syndiotactic dimer model compounds. As
shown in Figure 6, it is interesting that the isotactic and
syndiotactic peaks are reversed in their relative peak positions
for H-4 and H-7 in the dimer model. Thus, syndiotactic peak
resonates downfield for H-4, but upfield for H-7 relative to the
isotactic peak. This trend has been reproduced qualitatively by
the same for the two CH protons, H-11 and H-12; i.e., the
3
syndiotactic peak resonates upfield relative to the isotactic peak.
This trend has again been reproduced by the chemical shift
calculations. As shown in Figure 7, the observed CH carbon
peaks, C-4 and C-7, of syndiotactic dimer model compound
resonate downfield from the positions of isotactic peaks. Simi-
1
the H chemical shift calculations as shown in Figure 6. From
Table 3, it can be seen that the calculated syndiotactic chemical
shifts are almost the same between the H-4 and H-7, i.e., 4.52 and
4
.51 ppm, respectively. However, the isotactic chemical shifts are
larly, the observed CH carbon peak, C-11, of syndiotactic dimer
resonates downfield from the position of isotactic peak. How-
ever, no significant chemical shift difference is observed for the
3
significantly different between H-4 and H-7, i.e., 4.48 and 4.57
ppm, respectively. Among the four calculated chemical shifts of
9
251
dx.doi.org/10.1021/ma2018777 |Macromolecules 2011, 44, 9247–9253

Macromolecules
ARTICLE
isotactic peak, but the opposite trend is found for H-4 in the
dimer model compound. In PLA, the stereoregularity of the CH
proton was reported to be more sensitive to the lactic acid units
1
8
attached to the end with the hydroxyl group (“O-terminus”).
This corresponds to position 7 in the dimer model compound.
1
Thus, the observed H chemical shift of PLA should be compared
with the calculated chemical shift of H-7, rather than H-4. For the
calculated C peaks, C-4 and C-7, the syndiotactic peaks appear
1
13
Figure 8. H and C NMR spectra of copolymer of PLLA and PDLA in
13
1
13
CDCl
(
3
. (a) Expanded CH proton region in the H NMR spectrum.
b) Expanded CH carbon region in the C NMR spectrum. The
assignments were reported previously.
downfield relative to the isotactic peaks in the dimer model
compound. In PLA, the CH carbon was reported to be more
sensitive to the lactic acid unit attached to the end with the
13
13,18
18
carboxylic group (“C-terminus”), which corresponds to peak
C-4 in the dimer model compound. Thus, the diad level splitting
1
13
of H and C NMR spectra of the CH group of PLA can
be satisfactorily interpreted by the chemical shift calculation
performed here.
The agreement between observed and calculated shifts in PLA
means that the time-averaged local conformations predicted in
this work are applicable to PLA in solution, and the origin of
tacticity splitting in PLA is due to both time-averaged conforma-
tions and the chemical shifts of the conformations. At present, it
is still unclear why the width of the tacticity splitting of the CH
group in PLA is different for H and C. Further chemical shift
calculation at the tetrad level of PLA by taking into account the
time-averaged conformations for longer model compounds will
hopefully provide more information in the future.
1
13
Figure 9. Calculated H (left: (a) H-4, (b) H-7) and C (right: (a) C-4,
1
13
(
b) C-7) chemical shifts (in ppm) of dimer model compound (2) of
1 13
poly(lactic acid). (c) Observed H (left) and C (right) diad chemical
shifts of poly(lactic acid). All chemical shifts are shown relative to the
isotactic chemical shift. The blue stick corresponds to isotactic and the
pink syndiotactic.
CH carbon, C-12, between the two samples. As for the three
3
’
ACKNOWLEDGMENT
carbonyl carbons, C-2, C-5, and C-8, all syndiotactic peaks in the
dimer model compound are found to be upfield from the
isotactic peaks. Again, these trends have been reproduced by
the chemical shift calculations.
Thus, chemical shift calculations for isotactic and syndiotactic
diad model (2) of PLA coupled with the conformational energy
calculations can be used to predict time-averaged local confor-
mation and tacticity splitting in this PLA dimer. The chemical
shift is a good indicator that reflects the local conformations in
solution.
T.A. acknowledges support from Grant-in-Aid for Scientific
Research from Ministry of Education, Science, Culture and
Supports of Japan (23245045) and (21550112). The authors
would like to thank Mr. Masato Komiyama at Teijin Pharma Ltd.
for his support on the synthesis of the model compounds and
Dr. Masao Hirasaka at Teijin Ltd. for his support on discussion.
Mention of trade names or commercial products in this publica-
tion is solely for the purpose of providing specific information
and does not imply recommendation or endorsement by the U.S.
Department of Agriculture; USDA is an equal opportunity
provider and employer.
(5). Analysis of PLA Stereoregularity at the Diad Level on
the Basis of the Chemical Shift Calculations of the Diad
1
13
Model Compounds. H and C NMR have been used to
12À18
analyze the stereoregularity of PLA.
The tacticity assign-
’
REFERENCES
ments have been reported at the hexad level partly for the
(
1) Okihara, T.; Tsuji, M.; Kawaguchi, A.; Katayama, K. J. Macromol.
carbonyl carbon group, but the assignments of the CH group
3
Sci. Phys. 1991, B30 (1&2), 119–140.
(2) Ikada, Y.; Jamshidi, K.; Tsuji, H.; Hyon, S.-H. Macromolecules
1987, 20, 904–906.
(
(
(
have not been reported yet. As for the CH group, Figure 8 shows
1
13
the expanded H and C peaks of PLA in CDCl together with
3
13,18
the assignments at the tetrad level reported previously.
3) Tsuji, H.; Ikada, Y. Macromolecules 1993, 26, 6918–6926.
4) Tsuji, H.; Ikada, Y. Macromolecules 1992, 25, 5719–5723.
5) Li, L.; Zhong, Z.; de Jeu, W. H.; Dijkstra, P. J.; Feijen, J.
When the previous assignments of the tetrads in Figure 8 are
reduced to the diad level, an interesting trend can be seen. In the
1
H spectrum, the isotactic diad-centered tetrad peaks tend to
Macromolecules 2004, 37, 8641–8646.
(6) Tsuji, H.; Horii, F.; Nakagawa, M.; Ikada, Y.; Odani, H.;
resonate at lower field and the chemical shift range is relatively
large compared with the syndiotactic diad-centered tetrad peaks.
Kitamaru, R. Macromolecules 1992, 25, 4114–4118.
(7) He, Y.; Xu, Y.; Wei, J.; Fan, Z.; Li, S. Polymer 2008, 49, 5670–
1
13
Contrary to the H NMR case, in the C NMR spectrum the
syndiotactic diad-centered tetrad peaks resonate at lower field
and the chemical shift range is relatively large compared with the
isotactic diad-centered tetrad peaks. These trends are summarized in
Figure 9, together with the calculated and observed CH chemical
shifts of the dimer model compounds mentioned above.
The positions of the calculated and the observed chemical
shifts are shown relative to the isotactic peak. As noted above,
the calculated syndiotactic H-7 appears upfield relative to the
5
3
4
1
675.
(8) Tsuji, H.; Ikada, Y. Macromol. Chem. Phys. 1996, 197, 3483–
499.
(9) Yui, N.; Dijikstra, P. J.; Feijen, J. Macromol. Chem. 1990, 191,
81–488.
(10) Fukushima, K.; Kimura, Y. Macromol. Symp. 2005, 224,
33–143.
(11) Fukushima, K.; Furuhashi, Y.; Sogo, K.; Miura, S.; Kimura, Y.
Macromol. Biosci. 2005, 5, 21–29.
9
252
dx.doi.org/10.1021/ma2018777 |Macromolecules 2011, 44, 9247–9253

Macromolecules
ARTICLE
(
12) Bero, M.; Kasperczyk, J.; Jedlinski, Z. J. Makromol. Chem. 1990,
91, 2287–2296.
13) Kricheldorf, H. R.; Boettcher, C.; Tonnes, K. Polymer 1992,
3, 2817–2824.
1
3
(
(14) Kasperczyk, J. E. Macromolecules 1995, 28, 3937–3939.
(15) Thakur, K. A. M.; Kean, R. T.; Hall, E. S. Macromolecules 1997,
30, 2422–2428.
(
(
16) Kasperczyk, J. E. Polymer 1999, 40, 5455–5458.
17) Chisholm, M. H.; Iyer, S. S.; McCollum, D. G.; Pagel, M.;
Werner-Zwanziger, U. Macromolecules 1999, 32, 963–973.
18) Zell, M. T.; Padden, B. E.; Paterick, A. J.; Thakur, K. A. M.;
Kean, R. T.; Hillmyer, M. A.; Munson, E. J. Macromolecules 2002,
5, 7700–7707.
19) Tonelli, A. E. NMR Spectroscopy and polymer microstructure;
VCH Publishers: New York, 1989.
20) Matsuzaki, K.; Uryu, T.; Asakura, T. NMR spectroscopy and
stereoregularity of polymers; Japan Scientific Societies Press: Tokyo, 1996.
21) Tonelli, A. E. Polymers from the Inside Out; John Wiley & Sons,
Inc.: New York, 2001
22) Brant, D. A.; Tonelli, A. E.; Flory, P. J. Macromolecules 1969,
, 228–235.
23) Meaurio, E.; Zuza, E.; Lopez-Rodriguez, N.; Sarasua, J. R.
J. Phys. Chem. B 2006, 110, 5790–5800.
24) Carbone, P.; Ragazzi, M.; Tritto, I.; Boggioni, L.; Ferro, D. R.
Macromolecules 2003, 36, 891–899.
25) Boggioni, L.; Bertini, F.; Zannoni, G.; Tritto, I.; Carbone, P.;
Ragazzi, M.; Ferro, D. R. Macromolecules 2003, 36, 882–890.
26) Wu, W.; Li, W.; Wang, L.; Zhang, P.; Zhang, J. THEOCHEM
007, 816, 13–19.
(
3
(
(
(
(
2
(
(
(
(
2
(27) Kang, Y. K.; Byun, B. J. J. Phys. Chem. B 2008, 112, 9126–9134.
(28) Inai, Y.; Ogawa, H. Kobunshi Ronbunshu 2010, 67, 214–223.
(29) McAliley, J. H.; O’Brien, C. P.; Bruce, D. A. J. Phys. Chem. A
2
9
1
008, 112, 7244–7249.
30) Schafer, A.; Horn, H.; Ahlnichs, R. J. Chem. Phys. 1992,
7, 2571–2577.
31) Schafer, A.; Huber, C.; Ahlnichs, R. J. Chem. Phys. 1994,
00, 5829–5835.
(
(
(
(
(
(
(
32) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
33) London, F. J. Phys. Radium. 1937, 8, 397–409.
34) McWeeny, R. Phys. Rev. 1962, 126, 1028–1034.
35) Ditchfield, R. Mol. Phys. 1947, 27, 789–807.
36) Wolinski, K.; Hilton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990,
1
12, 8251–8260.
37) Cheeseman, J. R.; Trucks, G. W.; Keith, T. A.; Frisch, M. J.
J. Chem. Phys. 1996, 104, 5497–5509.
(
9
253
dx.doi.org/10.1021/ma2018777 |Macromolecules 2011, 44, 9247–9253