The aromatic diamine 2-(4-aminophenyl)ethylamine (4APEA) is a potential monomer for polymers and advanced materials. Here, 4APEA was produced by fermentation with using genetically engineered Escherichia coli [Masuo et al., Sci Rep, 2016]. Optimizing fed-batch cultures of this strain produced the highest reported yield to date of 4APEA (7.2%; 3.5 g/L vs. glucose) within 72 h. Appropriate aeration was important to maximize production and avoid unfavorable 4APEA degradation. Fermented 4APEA was purified from culture medium and polymerized with methylene diphenyldiisocyanate and hexamethylene diisocyanate to produce polyureas PU-1 and PU-2, respectively. The decomposition temperatures for 10% weight loss (Td10) of PU-1 and PU-2 were 276°C and 302°C, respectively, and were comparable with that of other thermostable aromatic polyureas. This study is the first to synthesize polyureas from the microbial aromatic diamine. Their excellent thermostability will be useful for the industrial production of heat-resistant polymer materials.

Bio-based polyureas (PUs) with main-chain furan rings were synthesized by the polyaddition of 2,5-bis(aminomethyl)furan with various diisocyanates, such as methylene diphenyl diisocyanate. Several PU’s were soluble in polar organic solvents, and were cast to form thermomechanically stable films with softening temperatures of over 100 °C. The furan rings of the PU main chains underwent a dynamic Diels-Alder (DA) reaction with bismaleimide (BMI) cross-linkers. While the mixed solution of PU and BMI did not show any apparent signs of reaction at room temperature, the DA reaction proceeded to form gels upon heating to 60 °C, which became a solution again by further heating to 80 °C (retro-DA reaction). The solution phase was maintained by rapid quenching from 80 °C to room temperature, while the gel was reformed upon slow cooling. The recovered gels exhibited self-healing properties. A scratch made by a hot knife at temperatures above 80 °C disappeared spontaneously. When two different gels were cut using a knife at room temperature, placed in contact with each other, and heated to 60 °C, they fused. The ability to control the DA/retro-DA reaction allowed gels of varying composition to heal.

Most of the state-of-the-art speech recognition systems use continuous-mixture hidden Markov models (CMHMMs) as acoustic models. On the other hand, it is well known that discrete hidden Markov model (DHMM) systems show poor performance because they are affected by quantization distortion. In this paper, we present an efficient acoustic modeling based on discrete distribution for large-vocabulary continuous speech recognition (LVCSR). In our previous work, we proposed the maximum a posteriori (MAP) estimation of discrete-mixture hidden Markov model (DMHMM) parameters and showed that the DMHMM system performed better in noisy conditions than the conventional CMHMM system. However, we conducted the recognition experiments on a read/speech task in which the vocabulary size was only 5k. In addition, the DMHMM was not effective in clean condition in that work. In this paper, we have developed a DMHMM-based LVCSR system and evaluated the system on a more difficult task under clean condition. In Japan, a large-scale spontaneous speech database 'Corpus of Spontaneous Japanese' has been used as the common evaluation database for spontaneous speech and we used it for our experiments. From the results, it was seen that the DMHMM system showed almost the same performance as the CMHMM system. Moreover, performance improvement could be achieved by a histogram equalization method. 