Dark dark chocolate

When I was in college, one of my closest friends told me that she only liked chocolate that was 80% or higher. I didn’t understand the concept at the time and was still a predominantly milk chocolate eater. I’ve since learned significantly more about the meaning of percentages and the virtues of dark chocolate.

Our chocolate creations have ranged from 70% to, at the highest, 85%. Well, we’ve tasted the ridiculously dark 100% bar from Dandelion and Endorfin’s 98% bar (2% vanilla). Personally, I think such high percentages taste more muddy than chocolatey. But, Richard is a huge fan. And more importantly, our neighbor, Jude, has personally requested a low-sugar bar to mimic the high percentage, bitter chocolate she knows and loves from her hometown of Barcelona. And given her current pregnancy and her self-proclaimed (and incredible) heightened taste buds, we’re excited to comply!

So, we set out to make a truly dark chocolate. We know it’s important to consider the cocoa butter content of our beans. And we have not measured the exact percentage of our Madagascar beans, but we’ll assume 50% for now. They seem pretty oily and our 85% was successful (though quite strong) when we tried previously.

We started, as usual, by measuring our initial cocoa bean batch (after sorting, before roasting). It came in at 1128 grams. We roasted in our beautiful Behmor, indicating 1 pound on the P2 program minus 2 minutes (as recommended by John Nanci). We took the beans (and nibs) through 4 passes of our homemade winnower, reducing the mass to 943 grams on the first pass (when we noticed lots of big shells), followed by 836 grams on the second pass (when we noticed that the nibs and shells had a lot of static energy), followed by 756 grams on the third pass (when it looked pretty good despite a few shells), followed by 722 grams on the fourth pass, which we deemed finished enough. In other words, we had a 64% yield on winnowing. Richard is still working hard on improving our winnowing process and tools!

We took some advice from the Industrial Chocolate Manufacture and Use book and immediately winnowed the beans, rather than waiting for them to cool. Then we tried a new step in our process: we put the nibs through a quick pre-grind in our new Vitamix 5200 Series Blender, when it turned the discrete nibs into smaller chunks with a slightly oily finish. We heated the stone wheels and reheated the beans, then started the melanger with the 722 grams of nib mush and 100 grams of sugar.


This resulted in an 87.5% dark chocolate with no added cocoa butter. We let it conche and refine for 24 hours, then tried the Chocolate Alchemist’s suggestion for tempering. We poured about a third of the chocolate onto plastic wrap and let it cool slowly in the oven, while the melanger continued conching. This created an effective “seed” chocolate. About an hour or so later, the seed had cooled, and we introduced it back into the warm liquor (at that point around 99 degrees). As the chocolate chunks mixed with the liquid, the temperature dropped significantly, and when we turned off the Premier Wonder Grinder, the temperature of the chocolate had just hit 90. We quickly doled it out into molds with our quick refrigerator pass to complete the tempering process.

And now our dark dark chocolate is ready for gifting!

What’s the highest percentage chocolate you’ve enjoyed eating? Any recommendations for low sugar, high percentage chocolate-making?


Eight-five percent

Last week, we finished another batch of delicious chocolate. This time, we tried four distinctly new ideas: a new roasting profile, a higher percentage cocoa, a longer refining time, and finally, a different temperature range for tempering. And the result? Smooth deliciousness!


Rather than our typical 400 degree hit, followed by 15-20 minutes at 250, we tried a roasting profile inspired by some comments Chloe made when we chatted at Dandelion a few weeks ago. This time, we let the initial 1249 grams of Madagascar cocoa beans roast at 225 F for 45 minutes. That’s the longest we’ve ever roasted beans, even for our first roasting test, when we came up with the Xtra Toasty chocolate! We quickly cooled the beans on our quartz board and let them sit there for about 15 minutes.

After passing the beans through mock 2 of Richard’s fancy winnower (see the beans in their hopper below) and a quick pass through with the hairdryer, we were left with 878 grams of nibs.



Now comes the second experiment. We added only 155 grams of further refined cane sugar. The result is an eighty-five percent bar. This diverges from the seventy percent bar that many consider the most optimum way to experience chocolate. Dandelion did an eighty-five percent bar last year with Rio Caribe (Venezuelan) beans, and we really liked it! Dandelion even has gone so far as to develop a 100% chocolate bar – that’s no sugar added. Their cafe manager, Jenna, writes about her experience with the 100% bar here.

Our final eighty-five percent bars are definitely intense with a lot of really interesting flavor notes. We’ll continue to experiment with higher percentages, taking our chocolate even closer to its roots!

Refining Time

This time, we let the Premier Wonder Grinder run for a full thirty-three hours, nine hours longer than we’ve ever done before. We added the refined sugar about nine hours into the refining process, to let the nibs refine on their own first. Since then, we’ve learned more thanks to comments from Robbie of Ritual Chocolate on a previous post. We’ll continue to refine our sugar refining process and will try adding it sooner next time! Either way, these thirty-three hours made the chocolate quite smooth and creamy. I think we’ll aim for these longer refining periods moving forward.

Tempering Temperatures

The final experiment has to do with tempering temperatures. We tried to follow Mr. Van Leer’s advice and initially raised the temperature of the chocolate to 105 degrees. We brought it quickly down to 84 on the quartz, table tempering the chocolate liquor. Then, we raised it up to 90 degrees in the microwave. The result was a failed knife test. Clearly, this is still the most challenging part of our chocolate-making.

We tried again, taking the temperature up to 112 in the microwave, then dropping it to 82 on the quartz table, then bringing it back up to 92 before pouring into the molds. We still had a small amount of bloom, but it looks better than previous batches.

root chocolate poured into molds

We’re improving and learning so much! Please let us know if you have any advice for us, particularly around tempering!

Just a quick note of thanks to all of our readers – please let us know if there’s anything in particular you’d like us to look into, investigate, or experiment with. We’re open to ideas and very invested in keeping this a genuine and honest source of information for chocolate makers and afficionados everywhere!

Sugar sugar!

Every time we make chocolate at home, we try a new experiment. Richard is a scientist, after all; this is in his DNA! Sometimes, we test different cocoa beans, like we did here. Other times, we test percentages of sugar. This time, we tried different varieties of sugar. And wow, we learned a lot!

With the same exact cocoa beans, percentages, roasting times and temperatures, and process, we created three batches. The only difference among the batches was the type of sugar. Our first batch, we’ll name it A, used regular bleached cane sugar. The second batch, let’s call it B, included raw sugar or turbinado. And the third batch, you guessed it – C, had Truvia, made from stevia.

We also tried a new method of cooling. Rather than plopping the finished chocolate onto the granite slab and separating it into bite sized pieces with our paint scraper, we created “kisses.” We filled Ziploc bags with the chocolate liquor after it was roasted, winnowed, ground, and tempered and squeezed it out onto the slab. We ended up with about 50 tiny kisses per batch (yes, we’re working with infinitesimal quantities at the moment).

The first major lesson we learned from this process was that stevia is really sweet. That seems obvious, since it’s sugar. However, if you taste it right before or right after cane sugar, the difference between the two is palpable. The difference between turbinado and cane sugar is less stark but still noticeable. And this is all before we put it in chocolate.

Next, during the grinding process, we learned that turbinado results in a drier mixture. So, even though we still refrained from adding any cocoa butter to the list of ingredients, the result was thicker and a little grainier than the other two samples.

With our new cooling method, we learned that the crystalization process happens more thoroughly when in direct contact with the granite slab. In other words, our kisses were well crystalized, harder, smoother, and more likely to have that traditional chocolate crack when we tried to break them apart, but only in the first centimeter or so from the granite. The parts of the kiss that didn’t touch the granite had the same partial crystalization that our previous chocolates had. We’re still working on optimizing this process to create the best possible texture and shelf life.

The fourth lesson takes us into scientific territory. Batch C with stevia provided a particularly strange tasting experience. Typically, if you were to put chocolate on your tongue, it would begin to melt as it warmed to the temperature of your mouth. However, the chocolate with stevia seems to cool as it melts. In other words, it melts colder than room temperature, rather than the other way around. Richard and my brother (a PhD student at Stanford) dug into the deeper meaning behind such a surprising phenomenon. They’ll have to follow up with their conclusions, because I don’t share in their scientific understanding. Another subject for another day!

Our final lesson was the most fun – taste testing the kisses and sharing them with friends and family. According to our many official tasters, the batch A with cane sugar was the clear winner. Batch B with turbinado came in a close second. And Batch C with stevia had only one fan of the many who tried the three batches. In fact, most people tasted one kiss and wanted nothing to do with it afterwards.

And as a bonus, we learned the opinion of a repeat taster, one who tasted our very first batch of chocolate from the beans we bought at the Grand Central Market in Los Angeles. She let us know that these three batches of chocolate’s texture, taste, and overall experience are leagues ahead of the first batch we ever made. Thank you, Helen, for being our first brand ambassador!

Update (August 14, 2014): For those of you curious about why stevia melts cold, Matt (my PhD of a brother) has figured it out:

So I found out why the chocolate with the Truvia causes a cooling effect in your mouth as you eat it. So, the sweetening agent in Truvia is stevia, but Truvia has other ingredients, like erythritol (erythritol is basically glycerol but with one extra carbon atom and hydroxyl group). Erythritol has a negative heat of solution, meaning that it takes more energy to dissolve the stuff than is released upon dissolution. So, as the erythritol dissolves, it takes up heat from its surroundings and the temperature decreases. So, if you drink something with erythritol in it, you do not get this effect since it is already in solution.


Cool, right? Mystery solved!