📄 chemistry

Density and molar volume of CaO–SiO2–FeO melts measured by electrostatic levitation aboard the International Space Station

Using the Electrostatic Levitation Furnace aboard the International Space Station, researchers measured the densities and molar volumes of CaO–SiO2–FeO melts under microgravity, revealing that density increases with FeO content while molar volume generally aligns with additive-model predictions despite slight deviations at higher FeO concentrations.

Original authors: Yusaku Seimiya, Masahito Watanabe, Takehiko Ishikawa, Chihiro Koyama, Yuki Watanabe

Published 2026-07-11✓ Author reviewed ⓘ

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Yusaku Seimiya, Masahito Watanabe, Takehiko Ishikawa, Chihiro Koyama, Yuki Watanabe

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a giant, floating soap bubble, but instead of soap and water, it's a glowing, molten ball of rock and metal. Now, imagine that bubble is floating in the middle of the International Space Station, where gravity is basically a ghost. That's exactly what happened in this study. Scientists from Japan took three different recipes of melted rock—mixtures of Calcium Oxide, Silicon Dioxide, and Iron Oxide—and sent them up to space to see how heavy they were when they were hot and squishy.

Why go to space? Well, on Earth, if you want to measure how dense a super-hot liquid is, you usually have to put it in a container. But hot metal loves to eat its container, or the container leaks into the metal, messing up the measurement. It's like trying to weigh a slice of pizza while it's still stuck to the box; the box adds weight and changes the flavor. In space, using a special "electrostatic levitation" trick, the scientists could hold the molten droplets in mid-air with invisible electric hands. No box, no container, just pure, floating liquid rock.

The team tested three specific "flavors" of this melt, labeled CS20F, CS60F, and CS80F. The numbers refer to how much Iron Oxide (FeO) was in the mix: 20%, 60%, and 80% by weight. The rest was a 50/50 split of the other two ingredients. They heated these droplets up using lasers until they were glowing hot, then let them cool down while snapping 60 pictures per second with a high-speed camera. By watching the shadow of the floating ball shrink as it cooled, they could calculate its volume, and since they knew the weight of the sample, they could figure out the density.

Here is what they found:
First, as the temperature went up, the density went down. It's like a balloon expanding when you blow hot air into it; the same amount of stuff takes up more space, so it becomes less dense. This happened in a straight line for all three mixes.
Second, at any specific temperature, the mix with more Iron Oxide was denser. The 80% Iron mix was the heaviest, and the 20% Iron mix was the lightest.
Third, they looked at something called "molar volume," which is basically how much space one "mole" of the mixture takes up. They found that when they added more iron (going from the 20% mix to the 60% mix), the space the mixture took up actually got smaller. But when they added even more iron (going from 60% to 80%), the space didn't change much at all. It was like squeezing a sponge: the first squeeze makes it compact, but the second squeeze barely does anything.

The scientists compared their space-measured data to old data from Earth and a simple math model called an "additive reference." This model assumes that if you mix two things, the total volume is just the sum of their parts, like stacking Lego bricks. The results showed that the real molten rock was mostly consistent with this simple Lego-block idea, though at the highest iron levels, the real melt took up a tiny bit more space than the Lego model predicted.

How sure are they? The paper states that the measurements had a maximum relative expanded uncertainty of 3.46%. This means if you took the number they got, the true value is likely within about 3.5% of that number. The biggest source of this tiny bit of uncertainty wasn't the floating ball itself, but the measurement of the reference sphere used to calibrate the camera. Because the volume calculation depends on the diameter of that reference sphere cubed (multiplied by itself three times), even a tiny wobble in measuring the reference sphere's size gets magnified.

So, what did they rule out? They didn't rule out the idea that containers mess up measurements; in fact, they confirmed that container-less methods are necessary for these high temperatures. They also didn't find that the volume changes wildly with iron content; instead, they found it drops sharply at first and then flattens out. They didn't claim this is a perfect, solved problem for all metallurgy, but rather that they provided a solid set of reference data for temperatures where Earth-based tools struggle.

In short, by floating molten rock in space, the team gave us a clearer picture of how these specific metal-slag mixes behave when they are hot and heavy, confirming that more iron generally means denser liquid, but the way the volume shrinks with iron content isn't a straight line—it's a curve that levels off. This data is now ready to help engineers and scientists model how metals flow and separate in high-temperature processes, using numbers measured without the interference of a crucible.

Technical Summary: Density and Molar Volume of CaO–SiO2–FeO Melts Measured by Electrostatic Levitation Aboard the International Space Station

Problem Statement
Reliable density data for molten slags are essential for analyzing fluid flow, heat and mass transport, phase separation, and interfacial phenomena in high-temperature metallurgical processes. The CaO–SiO2–FeO system represents a fundamental subsystem of iron-bearing metallurgical slags, necessitating accurate density data for process modeling and thermophysical property databases. Conventional contact methods, such as the maximum-bubble-pressure and Archimedean techniques, often suffer from contamination or reactions between the melt and supporting materials (crucibles, capillaries, bobs) at high temperatures. Consequently, there is a need for container-less measurement techniques capable of extending density measurements to higher temperatures without these interferences.

Methodology
This study utilized the Electrostatic Levitation Furnace (ISS-ELF) aboard the International Space Station (ISS) to measure the densities of CaO–SiO2–FeO melts under microgravity conditions.

Sample Preparation: Spherical samples (~2 mm diameter) with nominal FeO contents of 20, 60, and 80 mass% (designated CS20F, CS60F, and CS80F) were prepared. The balance of the composition consisted of CaO and SiO2 in a 1:1 mass ratio. Samples were melted in an aerodynamic levitation furnace on Earth to ensure sphericity before transport to the ISS.
Experimental Setup: Inside the ISS-ELF, samples were electrostatically levitated between three pairs of mutually perpendicular electrodes in an Argon atmosphere (2 × 10⁵ Pa). Heating was achieved using four 980 nm semiconductor lasers.
Data Acquisition: Sample temperatures were monitored via a monochromatic pyrometer (1.45–1.8 µm), and shadow images of the droplets were recorded at 60 fps using a CCD camera with ultraviolet backlighting.
Temperature Correction: For samples exhibiting recalescence (CS60F and CS80F), liquidus temperatures were calculated using FactSage 8.4 (FToxid database). The effective emissivity (approximately 0.8) was adjusted so that the indicated temperature at the recalescence inflection matched the calculated liquidus. This emissivity value was applied to the CS20F sample, which showed no recalescence and a glassy recovered surface.
Calculations: Droplet volume ( $V$ ) was determined by numerical integration of disk-shaped cross-sections derived from shadow image contours, assuming negligible gravitational deformation. Density ( $\rho$ ) was calculated as the ratio of the post-experiment sample mass ( $m$ ) to the volume ( $V$ ). Molar volumes were subsequently derived using nominal oxide compositions.
Uncertainty Analysis: Uncertainty was evaluated following the Guide to the Expression of Uncertainty in Measurement (GUM), considering measurement resolution and repeatability of mass, reference sphere diameter, and image factors.

Key Results

Temperature Dependence: For all compositions, density decreased approximately linearly with increasing temperature. The measurement ranges were 1331–1623 °C (CS80F), 1313–1934 °C (CS60F), and 1280–1942 °C (CS20F).
Composition Dependence: At a given temperature, density increased with increasing nominal FeO content.
Molar Volume Trends: Converted molar volumes at approximately 1400 °C showed a decrease from the 20 mass% FeO sample (CS20F) to the 60 mass% FeO sample (CS60F). Between the 60 mass% and 80 mass% FeO samples, the molar volume changed only slightly.
Comparison with Literature and Models: The measured densities aligned with the negative temperature coefficients and FeO dependence reported by Lee and Gaskell. The measured molar volumes were generally consistent with an additive reference model (treating the CaO–SiO2 fraction as a CaO·SiO2-equivalent component), though the experimental values were slightly larger at higher FeO contents.
Uncertainty: The maximum relative expanded uncertainty in density was 3.46% (coverage factor $k=2$ ). The largest contribution to uncertainty arose from the repeatability of the reference-sphere diameter measurement, as this parameter enters the volume calculation as a cubic term.

Significance and Claims
The authors state that these results provide density data for CaO–SiO2–FeO melts over a high-temperature range where conventional contact methods are difficult to apply. The data are presented as reference material intended to support metallurgical process modeling and the development of thermophysical property databases. The study demonstrates the capability of the ISS-ELF to perform stable, container-less density measurements of oxide melts, effectively mitigating issues related to crucible contamination and high-temperature limitations. The paper does not claim thermodynamic ideality for the system, noting that the additive reference used for comparison is a first-order volume additivity estimate.

More like this