To boil it down to basics, 8.5 is not Unicode compliant and in many fonts that have special characters including Chinese text, these characters are part of the extendend ASCII character set supported within Unicode for that font. Since 8.5 isn't Unicode compliant, issues will arise, especially when exporting, if your report uses special characters in the Unicode character set.
Upgrading to 9.0 will solve this issue as it is fully Unicode compliant.
The following is taken directly from the Crystal 8.5 Monthly Database and Export Drivers Hotfix PDF file for July 2003.
ADAPT00118085 Patch ID: 35059616
Description: When exporting a report that contains strings encoded in a particular character set (for example, German) to Text format on a system using a different code page than that set (for example, Japanese, or simplified Chinese), the export can enter into an infinite loop. The cause was that some characters were misinterpreted as the lead byte of a double byte character. This would result in the export to loop on the character following the misinterpreted one.
New Behavior: A different function is used to process the encoded strings so the export no longer loops indefinitely.
Known Limitations: This fix prevents the export from going into a loop. It does not correct the exported data. Because the print engine does not support Unicode, the export must assume the input is from the system default character set. Beginning with version 9, the Crystal print engine supports Unicode.