The paper describes the use of rotor-synchronized RFDR pulse trains on two nuclei for simultaneous heteronuclear and homonuclear polarization transfer. Such a sequence can be used to replace a heteronuclear polarization-transfer step (e.g., CP) and a consecutive homonuclear polarization transfer step by a single element that achieves both transfers. The new method is demonstrated using simulations and experimental data on SH3. At fast MAS or probably more likely for rf pulses that occupy a large part of the rotor cycle, heteronuclear and homonuclear transfer are almost identical while for slow MAS or rf pulses that are short compared to the rotor cycle, little transfer is observed.

In Figure 2, the CP and RFDR overlay spectrum is virtually invisible. Please improve the graphic representation. It would also be more interesting to see a 1D trace comparison of the HET-RFDR and the CP+RFDR spectrum and not two spectra at different mixing times. I think it would be nice to discuss the different transfer characteristics of the two sequences since there are marked differences in cross peak appearance.

Why is the transfer to the Ca in Fig. 3f negative? Is this a double-quantum type Hamiltonian that is generated here? Can you please comment in the paper whether the observed variations in the transfer efficiency are in agreement with expectations from the simulations/operator analysis?

Is there a reason why the rf-field amplitudes used in Figs. 4 and 5 are different? 83 vs 65 kHz? It would be much nicer to use the same rf field to make them directly comparable. I guess the simulation effort is not very high to change Fig. 5. I also think it would be nice to duplicate Fig. 5 in te SI for fast spinning so one can better understand the differences between the two cases in Fig. 4. 

This is an interesting and potentially important manuscript, demonstrating the utility of simultaneous RFDR pi pulse trains on 15N and 1H as a means of transferring polarization from 15N to 1H and also transferring polarization among 1H spins.  The authors show that long-range 1H-1H transfers are observed, which provide useful structural information.  With numerical simulations, they explore dependences on resonance offsets that are important especially at lower MAS frequenices.

This paper is certainly suitable for publication in MR.  My only recommendation is that the authors re-examine their choice of references in the Introduction.  It is worth noting that the first examples of homonuclear dipolar recoupling (by Meier and Earl for 1H-1H couplings and by Tycko and Dabbagh for 13C-13C couplings and quantitative distance measurements) are not cited.  The Ok paper is not about RFDR.  What is now called RFDR was originally introduced by Gullion and Vega (and called SEDRA).  The 2001 paper by Ishii analyzed finite-pulse effects in detail and showed that XY4 phase cycling produces an average homonuclear dipole-dipole Hamiltonian with the same operator form as in a non-spinning sample.