Olympus has special shear wave couplant (see link).  I've used Wood's metal, which has a very low melting point.  If you have a friend with a pulsed IR laser you can do away with couplant altogether.  But even with old (super viscous) honey, I don't remember having problems getting generating shear waves in the rock.  If you're interested in the laser approach have a look at the second link.

Hope this helps.  Cheers.

http://www.olympus-ims.com/en/applications/ultrasonic-couplant/

http://acoustics.mines.edu/preprints/PRE46618.pdf

Thanks John. Cheers.

for s-wave in rocks and metals, I have been using a honey based coupland they work just fine,

here is a small guide of others couplants

http://www.npl.co.uk/acoustics/ultrasonics/research/guide-on-acoustic-emission-sensor-couplants

Excellent. Thank you William.

You can use any couplant with high viscosity. As William recommended using honey based couplant would be a better solution (economical and attainable).

Refer - Ultrasonic Waves in Solid Media

           by Joseph L. Rose

You can use this device, that don't require couplant gel:

https://www.proceq.com/product/e-modulus-pundit-dry-point/