I just want to add new partitions to an existing table(with club column and indexes) in different tablespace.
And I asked some questions about this operation.
if you want to move your lob segment too then you must use your second alter statement (the one with clob). I don't know what kind of indexes you have on table so it is hard to say. Do you have local indexes or global or both? you can use "update indexes" clause in alter statement so your indexes will be remain valid but splitting will take much more time. how many data do you have in your max partition and how many rows will be split from max partition?
How carefully did you test the two options - I've run up a couple of sample in 10.2.0.4 and neither option does quite what you want to do.
The general principle of "split partition" is that if either of the two new partitions would be empty Oracle decides to rename the existing partitions and create one new empty partition, but if there is data for both new partitions Oracle creates and populates two partitions and drops the original. To see what it has done you need to look at the object_id and data_object_id from user_objects; and you also need to check the status from user_indexes and user_ind_partitions.
If you don't specify the LOB storage then the new (empty) lobs go into the current tablespace of the existing LOB. If you specify the old and new locations for the LOB partitions then the old lob content is copied (in this version of Oracle) and the PK index becomes invalid and the relevant partitions of the other index become invalid. You can "update indexes" with the split command - but then that still means you recreate the PK index.
You might like to test the split without specifying a tablespace for the OLD partitions, but specifying the tablespace for the new paritions - in a quick test this seemed to leave the indexes and index partitions valid, and simply rename the old partitions, while just creating new empty partitions.
(NB 10g is old - the split partition code on 12.2 where I reran the tests seemed to be cleverer and worked perfectly with the second version of your split partition code).