Electronic Supplementary Information

1. Polymerization with using n-butylamine as an initiator

A solution of N-(4-nitrophenoxycarbonyl)-γ-benzyl-L-glutamate 1 (0.20 g, 0.5 mmol) and n-butylamine (4.0 μL; 3.0 mg, 0.02 mmol) in N,N-dimethylacetamide (DMAc)(0.25 mL) was stirred at 60 ºC for 48 h under nitrogen. The reaction mixture was poured into ether (200 mL), and the resulting precipitate was collected by filtration with suction and dried under vacuum to obtain poly(BLG) a whitish brown solid in 78 % yield: 1H NMR (CDCl3): δ= 0.89 (br, 3H, -CH2-CH3), 1.33 (br, -CH2-CH3), 1.49 (br, 2H, -CH2-CH2-CH3), 1.57-2.86 (br, 4H, -CH2-CH2-C(=O)-O- and -CH2-CH2-C(=O)-O-), 3.21, (br, 2H, -NH-CH2-CH2-), 3.96 (br, 1H, >CH-CH2-CH2-), 5.05, (s, 2H, -O-CH2-C6H5), 7.27 (br, 5H, -O-CH2-C6H5).

Scheme S-1

2. 1H NMR spectrum of poly(γ-benzyl-L-glutamate) obtained by using p-(tert-butyl)phenylmethylamine

Figure S-1. 1H NMR spectrum of poly(γ-benzyl-L-glutamate) obtained by using p-(tert-butyl)phenylmethylamine.

3. MALDI TOF-Mass analysis

Matrix-assisted laser desorption / ionization time-of-flight mass spectrometry (MALDI-TOF MS) was carried out on a PerSeptive Biosystems Voyager DE Pro Bio Spectrometry workstation. 2-(4’-Hydroxyphenylazo)benzoic acid was used as matrix.

Poly(BLG) samples (1.5 mg) and 2-(4’-hydroxyphenylazo)benzoic acid (10 mg) were dissolved 1 mL of tetrahydrofuran, and stirred for 10 minutes. 2 μL of this solution was dropped to a sample plate and was dried under air. For ionization of the sample,337 nm nitrogen laser was irradiated. The resulting ionized polymers were detected in a positive mode with 20 kV for ion acceleration.

Poly(BLG) obtained by using 4-(tert-Butyl)phenylmethylamine (2) or n-butylamine were analyzed by MALDI-TOF mass spectrometry. Figure S-1a (in this Supplementary Information) is the spectrum of the polymer obtained by using 2 as an initiator. In the spectrum, two series of signals are observed: The series-A involves signals with m/z=1151, 1370, 1589, and 1808, and the series-B involves 1167, 1386, 1605, and 1824. In both of the series, the signals are regularly located with an interval of 219 Da, which corresponds to the formula weight of the repeating unit. Similarly, in the spectrum of the polymer obtained by using butyl amine as an initiator (Figure S-1b), two series of mass signals were observed (series A’: m/z=1062, 1281, 1500, 1719; series B’: 1078, 1297, 1517, 1736) and in both cases, the interval was 219 Da. The simple spectra with the regular intervals supported the polymerizations proceeded without any significant damage on the side chains.

Between the series-A and the series-A’, there was a gap of 90 Da. Similarly, between the series-B and the series-B’, there was also a gap of 90 Da. This gap agrees with the difference in molecular weight between the initiators, 2 and butylamine. Consequently, all of the signals would be attributable to the polymers having the amine-derived residue at the initiating end.

The final issue is the structure of X in the other terminal. Between the series-A and the series-B (for the polymers obtained by the 2-initiated polymerization), there is a gap of 16 Da. In both cases, the polymers would have the 2-derived structure at the initiating end, and thus there would be two kinds of terminating end structures, X1 and X2, of which formula weights were calculated to be 112 and 130, respectively. An analogous discussion on the series-A’ and the series-B’ leads to the conclusion that the polymers obtained by the butylamine-initiated polymerization would also have the same X1 and X2 as terminal structures. The most reasonable candidate for X1 would be a five-member cyclic amide-type one, which can be formed by intramolecular nucleophilic attach of the terminal amino group to the benzyl ester in the previous unit.[1] For X2, a candidate could be acyclic amino acid-type one, which may be formed by hydrolysis of the five-membered cyclic amide.

Figure S-2. MALDI-TOF Mass Spectra of the Polymers

Table S-1. Candidate Structures of Terminating End of the Polymers and the Corresponding Calculated Formula Weights.

4. 1H-NMR analysis of the reaction mixture

The reaction of 1in the presence of n-butylamine ([1]/[n-butylamine] = 50) in DMAc at 60 ºC was monitored with 1H NMR. After 3h, a small amount of reaction mixture was taken out and was dilutedwith CDCl3. The 1H NMR spectrum of this sample was shown in Figure S-3. At 4.4 ppm, a signal assignable to the methine proton of unconsumed 1 was observed. The signal at 3.8-4.2 ppm was assigned to the methine proton of poly(BLG). Besides these signals, there was a signal at 4.7 ppm, which was assigned to the methine proton of BLG-NCA. This assignment was confirmed by the 1HNMR analysis of the authentic BLG-NCA, which was prepared by the treatment of BLG with triphosgen and purified by recrystallization. The abovementioned results obtained by 1HNMR analyses of the polymerization confirmed that the ring-opening polymerization of BLG-NCA would be one of the major pathways to give poly(BLG).

Figure S-3.1H NMR analysis of the reaction of 1 in DMAc with using n-butylamine ([1]/[n-butylamine] = 50) at 3h. Solvent for the measurement= CDCl3.

5. Possible pathways involved in the present polymerization system

Scheme S-2

6. 1H NMR spectrum of poly(γ-benzyl-L-glutamate-co-ethylene glycol)

Figure S-4.1H NMR spectrum of poly(γ-benzyl-L-glutamate-co-ethylene glycol).