Amorphous Nanoclusters Promise Sharper MRI Scans at a Fraction of the Dose

Magnetic resonance imaging has long relied on a chemical compromise. Contrast agents sharpen the images that doctors depend on to find tumors, map blood vessels, and assess organ health, but the gadolinium chelates used in clinics worldwide deliver surprisingly modest signal boost per unit of metal. That weakness forces radiologists to inject relatively high doses, and it has kept a generation of materials scientists hunting for particles that brighten T1-weighted images far more efficiently. A team reporting in Materials Today Bio now describes a family of ultrasmall, amorphous paramagnetic nanoclusters that appears to clear that bar decisively, with longitudinal relaxivity values that dwarf commercial agents while remaining simple enough to manufacture at scale.

The work, led by Shilin Xiao, Mingfu Gong, and Dong Zhang at Army Medical University in Chongqing, centers on a deceptively simple observation: disordered matter holds water better than crystals do. Amorphous calcium carbonate, the same ephemeral mineral that shells and skeletons build before crystallizing, is riddled with pores and structural water, and its atoms lack the long-range order that locks metal ions into fixed coordination geometries. The researchers reasoned that if paramagnetic ions such as manganese or gadolinium could be trapped inside such a disordered matrix, nearly every one of them would end up surrounded by an unusually large number of water molecules, precisely the condition that magnetic resonance theory says maximizes T1 contrast.

To test the idea, the team mixed a polycarboxylic acid, typically polyacrylic acid, with a metal chloride salt in water, then added sodium carbonate and stirred for a single minute. The result was a burst of nanoclusters roughly two nanometers across, each a tangle of carbonate, metal ions, and polymer. X-ray diffraction confirmed the particles were amorphous, and the polymer coating gave them strongly negative surface charges and excellent colloidal stability. When dispersed on a clinical 3.0 tesla scanner, the manganese carbonate clusters showed a longitudinal relaxivity of 26.71 mM⁻¹s⁻¹ and the gadolinium carbonate clusters reached 38.64 mM⁻¹s⁻¹, roughly five to eight times higher than the clinical agents Gd-EOB-DTPA and Gd-DTPA measured in the same experiment.

The mechanism behind these numbers is a story of three relaxation parameters acting in concert. Solomon-Bloembergen-Morgan theory holds that T1 relaxivity rises with the number of water molecules directly coordinated to the paramagnetic center, with the number of hydrogen-bonded water molecules in the second coordination sphere, and with the rotational correlation time of the whole complex. Thermogravimetric analysis showed the nanoclusters carried water contents near 20 percent by mass, an extraordinary figure the authors attribute to dual hydration from both the amorphous lattice and the hygroscopic carboxyl groups of the polymer. Control experiments made the case convincing: manganese ions mixed with preformed amorphous calcium carbonate jumped to a relaxivity of 31.08 mM⁻¹s⁻¹, while the same ions mixed with crystalline calcium carbonate barely budged. The disorder itself, not the chemistry of the carbonate, is doing the heavy lifting.

Just as important is what the new particles do not do. Many high-relaxivity nanoparticles fail as T1 agents because their transverse relaxivity, which darkens tissue, rises even faster than their longitudinal relaxivity, pushing the r2/r1 ratio above 10 and turning images muddy. The optimized nanoclusters kept that ratio below 3, comfortably within the range expected of a clean T1 agent. The team further discovered that doping the clusters with non-paramagnetic ions such as magnesium, calcium, or zinc boosted relaxivity by roughly 40 percent. Because these ions carry no unpaired electrons, the effect cannot come from stronger magnetization; instead, water molecules hydrating the dopant ions sit close enough to the paramagnetic cores to feed the second-sphere relaxation pathway, an enhancement the authors argue outperforms conventional ligand-water interactions because of the shorter distances involved.

Equally notable is the synthetic economy. High-quality magnetic nanoparticles are conventionally made by thermal decomposition in boiling organic solvents at temperatures exceeding 250 degrees Celsius, followed by laborious ligand exchange to render them water-compatible. The new route runs at room temperature in water in about a minute, uses only polyacrylic acid, a metal chloride, and sodium carbonate, and tolerates variation in reaction time from one to 120 minutes without measurable loss of performance. A tenfold scale-up demonstration, using 10 millimoles of manganese, produced particles with a relaxivity of 24.73 mM⁻¹s⁻¹ at roughly 70 percent yield, evidence that the chemistry survives the transition from bench to batch production that kills so many nanoparticle platforms.

The in vivo results are where the work becomes genuinely striking. Amorphous manganese carbonate nanoclusters, injected at 25 micromoles of manganese per kilogram, lit up mouse livers with a contrast-to-noise ratio matching that of the clinical hepatocyte-specific agent Gd-EOB-DTPA at the same dose, and the enhancement persisted for hours rather than minutes. The team traced the liver specificity to SLC39A14, the transmembrane transporter that hepatocytes use to import manganese; cells engineered to lack the transporter took up far fewer particles, and gadolinium clusters without manganese showed no hepatic enhancement at all. In mice bearing hepatocellular carcinoma, tumors as small as one millimeter stood out clearly even at one-fifth of the standard imaging dose, a sensitivity that could matter enormously for early detection of one of the world’s deadliest cancers.

Clearance behavior adds to the appeal. Roughly half of the injected manganese clusters exited through the kidneys within days, and electron microscopy of the urine confirmed the particles left the body intact rather than dissolving into free metal ions. The blood half-lives were short, 1.4 minutes for distribution and 18.9 minutes for elimination of the manganese agent, which limits its use in long-duration angiography but sharply improves its safety margin. The gadolinium version, with slower clearance and a 127-minute elimination half-life, proved ideal as a blood-pool agent: at one-tenth the gadolinium dose of Gd-DTPA, it produced markedly clearer vascular reconstructions with cleaner background signal, addressing persistent concerns about gadolinium deposition in patients who undergo repeated contrast-enhanced scans.

Safety testing was thorough, if still preclinical. Cell viability remained above 80 percent across the full concentration range tested, hemolysis of mouse red blood cells stayed below 2 percent, and mice injected at four to five times the imaging dose showed normal blood chemistry and clean histology across six major organs at 28 days. The particles themselves proved shelf-stable, retaining their amorphous structure and relaxivity for more than a year of room-temperature storage. The authors are candid about the limits of their mechanistic account: without nuclear magnetic relaxation dispersion profiles, the precise values of coordinated and second-sphere water molecules remain inferred rather than measured, and attempts to build iron-based analogs failed for now. Still, the platform’s combination of one-step water-based synthesis, composition tunability, dual-mode liver and vascular imaging, and dose-sparing performance marks it as one of the more complete demonstrations yet that amorphous, rather than crystalline, nanoscale architecture may be the key to the next generation of MRI contrast agents.

Subject of Research: Ultrasmall amorphous paramagnetic nanoclusters as high-relaxivity MRI contrast agents

Article Title: Composition-tunable polycarboxylic acid-stabilized ultrasmall amorphous paramagnetic nanoclusters as multifunctional MRI contrast agents

Article References: Xiao, S., Sun, T., Zhang, L., Yang, Y., Yang, X., Liu, X., Zhao, Y., Zhang, Z., Zhang, W., Zhou, C., Gong, M., & Zhang, D. (2026). Composition-tunable polycarboxylic acid-stabilized ultrasmall amorphous paramagnetic nanoclusters as multifunctional MRI contrast agents. Materials Today Bio, 41, Article 103680. https://doi.org/10.1016/j.mtbio.2026.103680

Image Credits: AI Generated

DOI: 10.1016/j.mtbio.2026.103680

Keywords: MRI contrast agents, nanoclusters, amorphous materials, polyacrylic acid, manganese, gadolinium, relaxivity, hepatocyte-specific imaging, magnetic resonance angiography, SLC39A14, nanomedicine, co-precipitation