How to make a Contrast Matrix with 2 columns my file is about cell death n this experiment, the investigators treated cultured Arabidopsis protoplast cells with a mycotoxin that triggers programmed cell death. 

​> cont.matrix= makeContrasts(
+   Tsame = .T, 
+   Usame = .UT, 
+   Diff = ( .T- .UT) ,  levels = exp_design)
> fit2 = contrasts.fit(fit,cont.matrix)
> fit2 = eBayes(fit2)
> factor.vector=factor(c('T','T','T','T','U','U','U','U'))
> exp_design=model.matrix(~0 + factor.vector)
> colnames(exp_design)=c('.T','.UT' )
> fit=lmFit(eset,exp_design)
> exp_design

  .T .UT
1  1   0
2  1   0
3  1   0
4  1   0
5  0   1
6  0   1
7  0   1
8  0   1
attr(,"assign")
[1] 1 1
attr(,"contrasts")
attr(,"contrasts")$factor.vector
[1] "contr.treatment"

> cont.matrix= makeContrasts(
+   t= .T, 
+   u = .UT, 
+   Diff = ( .T- .UT) ,  levels = exp_design)
> fit2 = contrasts.fit(fit,cont.matrix)
> fit2 = eBayes(fit2)

If .T is your treated group, and .UT is your untreated group, then your Diff contrast will test whether the log-fold change upon treatment is significantly different from zero. Your Tsame and Usame contrasts don't make much sense; they are testing whether the average expression in each group is different from zero, which does not seem to be a scientifically relevant question (except, perhaps, for two-colour arrays).