1457
(a)
(b)
LL-lactide
DL-lactide
acyclic oligomer
DL-lactide
13C satellite
acyclic oligomer
acyclic oligomer
13C satelite
1
1
Figure 1. (a) H NMR spectrum of crude distillate obtained from oligo(lactic acid). (b) H homodecoupling spectrum of LL-lactide
crude distillate (Table 1, Entry 1).
Table 2. Reiterative production of L-lactide
Reiteration
yield of lactide decreased significantly (compare Entries 1, 6,
and 7). This unique DP dependence should be attributed, in part,
to decrease of the concentration of the hydroxy terminus in the
reaction mixture by condensation. With smaller (e.g., DP < 6)
oligo(lactic acid), the distillate was contaminated by lactic acid
oligomers. From Table 1, pyridinium triflate gave the best yield
at 180 °C, but the prolonged depolymerization did not improve
the yield of lactide very much.
In all cases in Table 1, pure lactide was obtained in good to
moderate yield. The purity of the distillate solid was determined
by 1H NMR (Figure 1). The lactide was free from other
impurities (<0.1%) (Figure 1a), such as lactic acid or oligo-
(lactic acid) of smaller DP, and the ratio of diastereomeric
lactides was determined to be >200:1 (Figure 1b). From this
value, the enantiomeric excess of the LL-lactide was calculated
to be >99.99%ee.4
1
2
3
4
5
Condensation step
OLA/mmola
436 423 424 417 410
Depolymerization step
Lactide/mmola
Yield based on OLA/%
112 101 99 105 108
25 24 23 25 20
Yield based on LA replenished/% ® 86 87 92 89
Hydrolysis step
LA replenished/mmola
aAs lactic acid unit.
117 114 114 121 ®
The reiterative lactide synthesis was realized. In this lactide
synthesis, depolymerization was carried out at 160 °C with Py-T
as a catalyst (Table 1, Entry 2). When Cl-Py-T was used
(Table 1, Entry 5), the residual poly(lactic acid) was contami-
nated with dilactilic acid, and with Py-T at 180 °C (Table 1,
Entry 1), elimination reaction produced alkene side products in
a significant amount. Thus, a mixture of lactic acid (90%) and
Py-T (1 mol %) was heated at 150 °C at 25 mmHg for 4 h to
afford oligo(lactic acid) of DP µ 9. The depolymerization at
160 °C at 1 mmHg afforded pure lactide in 25% yield (based on
the oligo(lactic acid)) leaving poly(lactic acid) of DP = 32 as a
distillation residue, which was hydrolyzed with aqueous lactic
acid at 130 °C for 15 h to regenerate monomeric lactic acid
(DP < 2). The amount of the added lactic acid is sufficient to
replenish the isolated lactide, and the amount of water is 1.1
equivalent to hydrolyze the poly(lactic acid) to monomeric lactic
acid. Then, the oligomerization-depolymerization-hydrolysis
sequence was reiterated. The lactide synthesis proceeded as
efficiently as the first run to give pure lactide, which exemplifies
that the catalyst is stable under the reaction conditions and the
side reactions such as elimination or formation of the ether
derivatives did not occur during depolymerization. The lactide
production cycle could be reiterated at least 5 times without loss
of the yield of pure lactide (Table 2). It should be mentioned
that, in this process, although the yield based on the oligo(lactic
acid) in each run is moderate, the yield calculated on the basis of
the replenished lactic acid after the second run reached 86-92%,
which shows the practical usefulness of the present procedure.
Nonetheless, in order to improve the efficiency of the
procedure, protection of the carboxylic acid group was examined
to suppress elongation of oligo(lactic acid) by condensation.
Inhibition of the condensation reaction during the depolymeri-
zation would increase the yield of the lactide. Eventually,
depolymerization of docosyl ester5 of oligo(lactic acid) of
DP µ 9,6 which was prepared by polycondensation of lactic acid
with docosanol (1/9 mol equiv) in the presence of Cl-Py-T
(1 mol % of lactic acid) at 140 °C at 25 mmHg for 7 h,7 afforded
99.5 wt % LL-lactide with >99.5%de and >99.99%ee in 80%
yield8 by similar depolymerization at 160 °C, 1 mmHg for 3 h.
Chem. Lett. 2012, 41, 1456-1458
© 2012 The Chemical Society of Japan
Ishijima, Yuuki
