You could look at some recent papers reporting mitochondrial genomes in plants.  I found this one, which is fairly recent: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4045975/

I would say there are two main sequencing technologies to look at.  Illumina HiSeq will provide large numbers of short reads with high accuracy.  PacBio SMRT technology will provide long reads.  

The first strategy, that has been popular in recent years, is to use Illumina paired-end sequences to construct a fragmented assembly, and then use either Illumina mate-pair reads or PacBio long reads to construct the scaffold that links the fragments together. This strategy is rooted in the view that Illumina reads are the most accurate.

For mitochondrial genomes, there may be a strong AT bias, whereas Illumina reads have tended to a GC bias, so you must take care of this.

Recently I think PacBio has been seen as more viable as a primary technology for small genome assembly, so an alternative strategy is to do your sequencing on PacBio and see how the genome assembles.  There is an option then to also capture some Illumina sequences, and use them in a pre-assembly error-correcting step to improve the accuracy of the PacBio reads.

Thanks Dr. David

I don't have experience in plant mt-genomes but  I have generated some animal mt-genome sequences using next-gen techniques. I used both 454 and Illumina and got consistent results in general. Even, I tried both LA-PCR enriched and PCR-free techniques to generate mt-genome sequences. I could generate the mt-genome sequences in both the methods. However, I always had a problem in repetitive regions. My opinion is if you have a good reference sequence to help you in assembly, any platform should be okay for you.

Best,

Mega

You can capture enrichment or use miseq library building.

There are several methods you could opt: 1) Illumina TruSeq amplicon sequencing approach, I don't know how big the mtDNA is for plant genome, no more than 20kb typically for mammalian mtDNA genome.  Through designing tens of Primer pairs covering the whole mtDNA genome, the subsequent sequencing will be quite compatible with illumina sequencing platforms, as well as supported with multiple sample pooling together; 2)  droplet PCR and digital PCR; 3) PacBio long-read technology;

Thanks to David, James Steward, Meganathan, Zhang,and Yanfeng Zhange for your kind answers and clarification to our (Naragarja Reddy's) question!

Thanks!