Espresso Centrifuge

Still trying to learn more about the filters to use in espresso filtration for refractometer measurements I emailed Wybo Dekker, a friend who has conducted chemical analysis all his life until his recent retirement. For many years he had his own consultancy firm. We have both been board members of the Dutch language TeX user group NTG and when meetings were held in his monumental home built long ago on a dike along a small river, I was always impressed with the ultramodern chemistry tools here and there in this beautifully restored old house.

Wybo could not tell me much about filters but he did have a brilliant offer:

"No, I was never seriously involved in these little filters so I can't tell you anything about them. As a matter of fact I hate filters --- they tend to get clogged on the most inconvenient moments, especially if for some reason you want to filter out relatively small particles. 
That's why I prefer to use an excellent centrifuge. I've got one here -- up to 4000 rotations per minute, electronically controllable, programmable including rates of accelleration and slowing down to avoid any turbulence during those phases. 
I have no clue as to what you're trying to accomplish with coffee, but if it is of any use to you you're welcome to have it. Mind you, it's a heavy thing measuring 40x45x50cm."

I was quite thrilled of course and in a hurry to pick up this gem. Until I had time to do so, I read the PDF of the manual which was available online and I watched some online video's about centrifugal separation. A machine like that needs careful treatment because even with relative small amounts of fluid rotating at thousands of RPM, the G-forces can be substantial and an imbalance can cause havoc.

So here it is, the Espresso Centrifuge next to the Londinium I:

Hettich Rotanta /P Centrifuge

On the back it has what looks like an RS232 port so who knows what I can yet discover about its remote control options. Wybo also gave me a number of little glass tubes to load the espresso fluids to separate from the tinest floating coffee grind particles, and a couple of very smart volumetric pipettes which can suck up and deliver a very precise volume of fluid, to the thousandth of a milliliter. He rarely used those but I think they look rather geeky so they were a must-have.

Time to brew some espresso! I used 20g of the Panama Los Lajones from Graziano Cruz which I roasted two weeks ago. Got 30.3g extraction:

30.3g of espresso from 20g grinds on the Londinium I

Next I weighed off 7g of the (stirred) espresso in a test tube and also 7g of water in a matching test tube:

Two identical test tubes, both filled with 7g, one water, one espresso

I placed these on opposing sides in their containers on the rotor, so the centrifuge will not have any imbalance:

Coffee and water of equal weight in matching seats, ready for the ride

Setting breaking force to 3 (out of a scale from 0=free running to 9=full brakes) and target speed of 3000 RPM, no pre-set running time:

Break force 3, rotation speed 3000 per minute, ready to start

The machine revved up slowly, gradually accumulating RPM speed and the digital dials showed actual speed and time since start.  After 4 minutes, at the STOP button the machine slowly calmed down and all signs blinked when it was safe to open and take out the sample.

I measured the sample straight from the test tube, using a new pipette onto the calibrated refractometer and read 9.6 % TDS:

9.6% TDS on sample out of centrifuge

Then it got a little more exciting still. I took a new sample out of the same espresso cup that had yielded the sample that I just measured, using a new syringe and a new VST filter and used a new pipette to put drops on the (cleaned) refractometer:

Yay! Filtered sample out of espresso source gives identical reading

This is nice. The filtered sample gives the exact same result on the refractometer, which indicates for now that the centifuge method of separation is effective enough to work with.

I will need to do more testing and I will specifically be trying to find out the best rotation speed and the optimal duration of the centrifugal separation. There was some sediment visible in the bottom tip of the test tube but I think the fluid could be separated even better. And then I could save this sequence into a numbered program, including the time to reach max RPM and the time to slow down to the full stop. I will then be able to reproduce the exact same procedure every time.

With this machine I can skip filtration if I want to.

Thank you Wybo for this beautiful device!