Do I have to change my Locale or database Encoding to enable LOWER() ?

Hello Hackers,
I store some information of manufacturing materials in a table, and in this table there's a column storing material's name. These names are mixed with Chinese characters and English characters.
To convert all English characters in material names into lower cases, I wrote a query: SELECT LOWER("ChineseDesc") FROM "Materials";

Then Postgres reported an error:
ERROR: invalid multibyte character for locale
HINT: The server's LC_CTYPE locale is probably incompatible with the database encoding.

After some googling I found out it's caused by the settings of LC_CTYPE & Encoding of the database.

Here's my db's settings:
My LC_CTYPE is set to Chinese_People's Republic of China.936 and I'm using UNICODE as this database's encoding.

But still I'm puzzled:
Which encoding to choose for my database or do I have to change my Locale setting? I prefer to stay with UNICODE, because I don't want to see my data clearly in PGAdmin 3 instead of *&^%$#@ .

Any suggestions will be appreicated!
Thanks in advance,

Henry

------------------------------------------
弄懂了PostgreSQL, 就知道了什么才是共产主义