– Kintsugi

This is the second article in my Kintsugi research series. For the overview, visit Kintsugi 101. Read on for instructions on how to join broken ceramic pieces together.

For my first project repairing a cracked bowl, I am using the following recipe for mugi-urushi:

  • 4 parts mochi rice flour
  • 1 part white wheat flour
  • water
  • seshime urushi from Mejiro

I mixed the flour and water to create a dough, then mixed 1:1 with urushi to create a paste. I applied this paste to the two halves of the bowl with a plastic spatula and squished them together without a waiting period.

Although Kintsugi Oxford recommends a wait period of 1 hour before pressing pieces together, does not. I decided to follow my gut here and squish them together right away, since my previous attempt at sabi started hardening within the first 10 minutes.

For the undercoat (Shita-nuri), I’m pretty sure I bought the wrong type of lacquer. Seshime Pure Lacquer dries light brown, and it’s almost translucent, whereas Kintsugi Oxford and show a black opaque lacquer. Plus, I found some sites referencing seshime urushi as a wood stain, which explains why it’s so thin. The buildable layers would look great on wood.

Even if seshime is the wrong type of lacquer, it still works to stick things together. So, after mixing and squishing, I wiped the excess mugi-urushi off with a spatula, positioned the cracks so they were seamless, and rubber banded the whole thing together.

If I bought a fancy bamboo spatula, this would look less crazy. But I’m being minimal with my materials until I get better!

Right after application, I wiped off the excess with a spatula and let it sit for 3 hours in the curing box. This image looks a little wild because of all the spatula marks, but they come off easily with a little sandpaper and water.

After 3 hours, I sanded off the excess with wet sandpaper and used acetone to clean up the edges. I was concerned about chipping and breakage with the rice flour based on’s opinion that rice flour is less malleable and more brittle than wheat flour. However, rice glue feels more traditional, and I couldn’t stop thinking about an old TV special I saw about mashing mochi rice to create a glue, similar to this process. So, I gave it a shot, and so far it hasn’t broken.

Note the nice fill of a piece of chipped glaze

The bottom had another piece of chipped glaze that the mugi-urushi filled in nicely, despite my best efforts to sand it away. 🙂

This portion chipped and crumbled away during sanding, but that’s OK because I will fill it in with sabi during the next stage

Overall I’m very happy with this mugi-urushi recipe for this type of crack repair. The bowl is very lightweight, so that might have something to do with it. In the future if I work on a more heavyweight ceramic repair, I’ll probably increase the wheat flour ratio.

After 14 days of curing, I’ll continue to the sabi phase!


It’s been almost half a year since I left this piece in the muro. It’s probably cured by now right? 😅

I was concerned about using rice flour as the primary binder instead of wheat flour. Since this bowl was at least 5mm thick, I was worried that the rice flour wouldn’t be strong enough to keep the pieces in place under pressure. So, I tried to use my strength 💪 to pull the bowl apart at the seams. Not my full strength, but I definitely went beyond the limits of the force a bowl would endure in daily life. The repaired bowl felt super strong, as if there was never a crack in the first place. I imagine it will only get stronger the more layers of urushi I apply in the next phase (Painting Urushi Lacquer).

Maybe a long cure time was accidentally the best course of action?

Research Notes

I’m interested in what the “standard” mugi urushi recipe is, so I’ve turned to Japanese articles and videos. Urushi-san, Tokyu Hands and Kintsugi Library recommend applying plain urushi to the cracked pieces before combining them. I didn’t do this in my first round!

Kintunagi of Ishikawa Prefecture and Urushi-san recommend wheat flour for the mugi-urushi mix.

Kintsugi Library’s tutorial seems much different from the rest in that the mugi urushi is thinner and the wait time before pressing the pieces together is 3-4 hours.1 After translating Step 06 (Make Adhesive with lacquer), looks like they use wheat flour too. Perhaps the liquid consistency can be explained by the amount of water added. After flour and water are kneaded, urushi is added at a 1:1 ratio.

In Step 07 (Brush Preparation), they oil the brush and arrange the fibers into a fine point with saran wrap. Next, they apply mugi-urushi to the cracks in a very thin layer, taking only small amounts at a time.

In Step 08, they wait until the mugi-urushi is tacky before pressing the pieces together. In winter (cold weather), this could take 3-4 hours.

I’d like to try this technique because it’s a lot cleaner looking and allows you to skip the acetone cleaning stage. However, the bowl in this tutorial is very thin and delicate, but my next bowl is super thick and heavy. I wonder if the wait period increases or decreases the tensile strength? What about the thickness of the adhesive applied? The absence of the acetone step?


This paper suggests that rice glue has poor tensile strength over time and that higher protein starches are better. Wheat is a pretty high protein flour.

This study shows that flour & water outperforms Elmer’s glue. This study showed that wheat glue on wood planks is very stable, but at temperatures over 105ºC the bond weakened. They attributed this to the thermal modification of the wood at that temperature. So as long as my ceramics will not change shape at high temperature, I think wheat glue will be fine.

All English language resources I found favor wheat flour. On to Japanese resources!

This blog explains that wheat flour has gluten, unlike rice glue. When processing rice glue in the traditional way, one must use fresh cooked rice and smash it to a paste, so my rice flour method is likely insufficient. This is backed up by the author’s research on starch types and how rice starch must be heated in order to gelatinize, while wheat flour can just be mixed with water at room temperature.

Duh! When I made strawberry daifuku, I had to microwave and stir the rice flour mixture to create mochi. Since I didn’t do this with the blue bowl repair, the adhesive strength must come from the wheat flour I mixed in. Oops! 😅

This article explains that all natural starch adhesives require gelatinization to be effective. Gluten, on the other hand, just requires water, and used to be used in industrial leather and wood manufacturing. This makes me think that using vital wheat gluten, much like seitan production, would be a better material than plain wheat flour.

From cooking homemade seitan, I know that the bowl is impossible to clean afterward even when soaking with water—unless you pre-grease the bowl. Compared to cleaning up from making homemade mochi, I think wheat gluten is the more powerful adhesive. Another observation from cooking: seitan dough seems to dry faster than mochi dough.

2022-03-13 Update The vital wheat gluten experiment was a big failure. The pieces did not fully cure even after over a week, just kind of fell apart. I’ll continue to experiment with wheat flour.

This article by a professor at Tokyo University of Agriculture explains that urushi lacquer was the first adhesive in Japanese history, used as an adhesive for bows and canoes. In the presence of oxygen, the sap of the urushi lacquer tree polymerizes and hardens.

This Tape Museum article explains that starch glues are used for creating shoji screens because it’s easy to remove the paper by moistening the glue. Well, this is a good reason to not use plain rice or wheat flour for kintsugi.

This author prefers rice glue made from scratch over wheat glue. The application was gluing wood planks together.

Apparently buddha statues use an adhesive of 45% rice flour, 45% urushi, and 10% wheat flour.

Overall it looks like there isn’t a clear answer. If prepared properly and mixed with urushi, it looks like either wheat or rice would work. Since I’m too lazy to mash my rice by hand, I’m going to use wheat gluten and wheat flour for my next experiment. Seitan glue! 😈

TLDR: Don’t use mochiko rice flour paste because it won’t gelatinize without heat!

  1. 01【割れた金継ぎ】お茶碗の直し方~接着まで by Kintsugi Library, Youtube

Here are all the notes in this garden, along with their links, visualized as a graph.