Tian Xie, principal research manager at Microsoft Research AI for Science
Photo: Courtesy of Tian Xie via his official personal website; photographer not stated · Publisher-directed editorial display; source copyright retained

FigureAsia 35 Under 35 · Science

Tian Xie

Age 32 · Generative materials discovery · China / United States

Led the development of MatterGen, a generative model for designing inorganic materials to specified properties.

Approximate age at the edition eligibility date
32
Field
Artificial intelligence for science
Country or region
China / United States
FigureAsia U35 Assessment
95.2 / 100

Career and documented record

Materials discovery has traditionally meant searching a database or screening candidates assembled by human intuition. Tian Xie led the development of MatterGen, published in Nature in 2025, which instead generates inorganic crystal structures conditioned on desired properties.

The model was trained on hundreds of thousands of structures and evaluated computationally across multiple property targets. The team also synthesised selected candidates, an essential step in showing that generated structures were more than plausible coordinates on a screen.

MatterGen does not automate materials science or guarantee manufacturable compounds. Its importance is methodological: it reframes inverse design as generation, joins machine learning to experimental validation and gives researchers a new starting point for navigating a vast chemical space.

Why Tian Xie is on the list

Xie pairs clear technical leadership with one of the period's most consequential AI-for-science publications. MatterGen is valuable not because it promises infinite discovery, but because it turns a difficult inverse problem into a tractable, testable workflow and then subjects selected outputs to experiment.

The 2025–26 record

MatterGen in Nature

Led a property-conditioned generative model for inorganic materials design.

Experimental follow-through

The research included synthesis and characterisation of selected generated candidates.

Open research platform

Extended generative modelling from structure prediction toward inverse materials design.

The work in its field

Generative models in materials science are judged by physical validity and experimental follow-through, not visual plausibility. MatterGen's strongest contribution is the property-conditioned workflow and its bridge to synthesis.

Assessment breakdown

95.2out of 100

01

Substantive 2025–2026 contribution

18.8 / 20

Led a property-conditioned generative model for inorganic materials design.

02

Verified scientific impact

14.1 / 15

Publication in Nature, experimental validation and rapid field uptake make MatterGen a reference point for generative materials research.

03

Originality and distinction

9.5 / 10

The distinction lies in generating crystal structures directly from target properties instead of ranking only known candidates.

04

Field influence

9.6 / 10

Within artificial intelligence for science, the work matters because it shifts a live question in generative materials discovery rather than merely attracting attention.

05

Individual agency

9.6 / 10

Microsoft Research identifies Xie as the scientist who led MatterGen's development, within a large multidisciplinary team.

06

Durability and trajectory

4.8 / 5

As Principal Research Manager, AI for Science at Microsoft Research, Xie has a platform to carry the work into its next stage.

07

Asian significance and global relevance

4.8 / 5

Chinese scientist whose education and research career connect China with the United States.

08

Evidential validity and reproducibility

7.7 / 8

Computational evaluation is reinforced by experimental synthesis; the profile does not generalise beyond the tested materials and properties.

09

Advance in scientific knowledge

6.7 / 7

The project shows that generative modelling can address inverse design under explicit physical-property constraints.

10

Translational or methodological utility

4.8 / 5

MatterGen can narrow candidate spaces and redirect laboratory effort toward compounds with specified performance targets.

11

Responsible research stewardship

4.8 / 5

The assessment credits the open scientific workflow while retaining the uncertainty between generated structure and deployable material.

Evidence and attribution

Material claims on this page are supported by the edition’s evidence record. FigureAsia tests age, identity, role, result and individual attribution before publication. Public profiles present the reported record; supporting documentation is retained for accuracy review and corrections.

Achievement records
3
Assessment window
2025–26
Editorial status
Included in the 2026 FigureAsia 35 Under 35 edition

Rights and credit

The portrait is published under the rights basis recorded for this edition. Third-party ownership and reuse restrictions remain in force.

Publication status
Published under a documented rights basis
Credit
Courtesy of Tian Xie via his official personal website; photographer not stated
Licence
Publisher-directed editorial display; source copyright retained
Portrait source and credit