RaspAssistant as an Agentic Framework for Automating Adsorption Simulation Research in Porous Materials

Marko Petković; Susanne Gooijer; José Manuel Vicent-Luna; Vlado Menkovski; Sofía Calero

cs.AIOct 20, 2026

RaspAssistant as an Agentic Framework for Automating Adsorption Simulation Research in Porous Materials

Authors: Marko Petković, Susanne Gooijer, José Manuel Vicent-Luna, Vlado Menkovski, Sofía Calero

Organizations: Department of Applied Physics and Science Education, Eindhoven University of Technology, Eindhoven, The Netherlands · Eindhoven Artificial Intelligence Systems Institute (EAISI), Eindhoven University of Technology, Eindhoven, The Netherlands · Department of Mathematics and Computer Science, Eindhoven University of Technology, Eindhoven, The Netherlands

Abstract

Porous-material adsorption studies require coordinated structure preparation, force-field selection, simulation, validation, and interpretation, making end-to-end automation difficult. We test whether RaspAssistant, a hierarchical multi-agent framework, can carry an adsorption study from force-field preparation through validation and downstream separation analysis. Across twelve multi-agent workflow-reproduction runs spanning four adsorption and separation benchmarks, overall scores ranged from 8.50 to 10.00 on a rubric evaluating workflow correctness, simulation validity, and analysis consistency. In a separate eight-case force-field extraction benchmark, the mean weighted score was 0.967, combining parameter agreement and RASPA readiness. In open-ended studies, the system diagnosed and refined an acetylene force field and, after validating a CO $_2$ force field against MFI experiments, screened 197 pure-silica zeolites that passed automated input checks. These results demonstrate autonomous workflow execution after task specification and input provision: RaspAssistant coordinates, validates, adapts, and documents extended adsorption-simulation workflows while exposing model sensitivity and methodological limitations for expert review.

Figures & tables

Figure 1: Agent architecture of RaspAssistant. Orange nodes denote higher-level agents responsible for planning, coordination, and verification, while blue nodes denote lower-level ReAct agents that alternate reasoning with tool actions. The red dashed subsystem supports literature search, force field extraction, adaptation, and parameter file generation. The blue dashed subsystem supports simulation setup, execution, analysis, plotting, and force field fitting for workflow reproduction. Shared databases, examples, memories, and tools occupy the bottom layer. Agent roles are listed in Table S5 .

Benchmark	Run	Overall (/10)	Elapsed (d)	Tokens (M)	Cost (US$)
TAMOF-1	1	10.00	8.8	12	15
	2	9.60	2.4	12	14
	3	9.60	8.8	7	8
Acetylene	1	9.05	60.0 *	21	17
	2	9.60	5.0	14	15
	3	8.73	36.3	11	11

Table 1: Scores and approximate resource use for the twelve multi-agent workflow-reproduction runs.

Method	IoU	Readiness	$S_{\mathrm{FF}}$	Cost ($)
CO 2 EPM2 41	0.95	100%	0.970	0.19
CO 2 in zeolites 42	1.00	88%	0.952	0.23
CO 2 in zeolites 43	0.99	92%	0.962	0.34
CO 2 , N 2 , O 2 , Ar in zeolites 44	0.98	100%	0.988	0.24
TraPPE-zeo 45	1.00	100%	1.000	0.15
C x H x , CO 2 , N 2 , O 2 in zeolites 46	1.00	100%	1.000	0.65

Table 2: Force-field extraction performance and weighted score by source.

Figure 2: Schematic overview of Case Study 1. Here, baseline validation means validation of the supplied initial force field. The workflow includes detection of the poor C 2 H 2 isotherm in Fe-MOF-74, convergence checks, force field fitting, validation of the fitted model, and subsequent mixture adsorption and adsorption site analysis.

Figure 3: C 2 H 2 adsorption isotherms in Fe-MOF-74. “Baseline simulation” uses the initial force field. “Revalidation simulation” uses the fitted force field. The plot was generated by RaspAssistant. The isotherm from Ref. 38 was added manually for comparison.

Model	Mean Fe–C distance (Å)
Initial force field	3.12
RaspAssistant fitted force field	2.63
Literature force field	2.60
Experiment	2.50

Table 3: Comparison of the arithmetic mean of the two Fe–C contacts for C 2 H 2 adsorbed in Fe-MOF-74.

		Agent		Literature
Interaction Pair	Atom Types	$\epsilon$ (K)	$\sigma$ (Å)	$\epsilon$ (K)	$\sigma$ (Å)
Metal site	Fe – C ${}_{\mathrm{C_{2}H_{2}}}$	680.0	2.355	38.92	2.878
Framework oxygen	O1 – C ${}_{\mathrm{C_{2}H_{2}}}$	10.0	2.900	79.23	2.870
	O2 – C ${}_{\mathrm{C_{2}H_{2}}}$	20.0	3.050	79.23	2.870
	O3 – C ${}_{\mathrm{C_{2}H_{2}}}$	20.0	3.050	79.23	2.870
Framework carbon	C1/2/3/4 – C ${}_{\mathrm{C_{2}H_{2}}}$	—	—	78.96	3.055

Table 4: Comparison of fitted Lennard–Jones parameters for acetylene–framework interactions in Fe-MOF-74. Literature parameters are taken from Ref. 38 .

Figure 4: Comparison of C 2 H 2 /CH 4 (a) and C 2 H 2 /CO 2 (b) mixture selectivities from the fitted Fe-MOF-74 and retained IRMOF-1 models with Ref. 38 . RaspAssistant generated the plot. Reference data were added manually.

Figure 5: Normalized minimum periodic interaction-site-to-Fe distance distributions in Fe-MOF-74 for equimolar C 2 H 2 /CH 4 and C 2 H 2 /CO 2 mixtures at 298 K and 1000 kPa. The sharp short distance peak for C 2 H 2 indicates preferential localization near the open Fe sites, whereas CO 2 and CH 4 are mainly distributed farther away in the pore space.

Figure 6: Schematic overview of the main steps taken by RaspAssistant in Case Study 2. The workflow includes preparation of literature-based force-field candidates, validation against MFI CO 2 adsorption experiments, selection of the best-performing force field, 298 K screening of the pure-silica zeolite dataset, and post-processing through ranking and descriptor analysis.

Figure 7: Validation of the candidate force fields against the provided CO 2 adsorption isotherm in MFI at 303 K. The TraPPE_Silica_2016 candidate combines TraPPE CO 2 with a Vujic-derived silica-zeolite model and gives the best agreement. Open markers flag block-drift outliers. These points are retained in the error calculations (see the SI).

Figure 8: Distribution of interpolated CO 2 uptake at 1, 10, and 100 kPa and of the derived 1–100 kPa working capacity in Case Study 2.

Figure 9: Relationship between framework density and simulated CO 2 uptake at 1 kPa for the screened pure-silica zeolites. Highlighted topologies correspond to positive outliers relative to the overall linear trend, indicating unusually high uptake for their framework density.

Run	Workflow	Simulations	Analysis	Overall
Run 1	10.00	10.00	10.00	10.00
Run 2	10.00	9.00	10.00	9.60
Run 3	10.00	9.25	9.50	9.60
Multi-agent average	10.00	9.42	9.83	9.73
Single-agent baseline	9.13	9.13	8.75	9.05

Table S1 : Multi-agent RaspAssistant performance and the single-agent baseline on Study 1.

Run	Workflow	Simulations	Analysis	Overall
Run 1	9.50	8.13	10.00	9.05
Run 2	10.00	9.00	10.00	9.60
Run 3	8.88	8.56	8.75	8.73
Multi-agent average	9.46	8.56	9.58	9.13
Single-agent baseline	7.50	8.56	8.13	8.05

Table S2 : Multi-agent RaspAssistant performance and the single-agent baseline on Study 2.

Run	Workflow	Simulations	Analysis	Overall
Run 1	8.69	8.75	8.00	8.58
Run 2	9.50	9.75	9.50	9.60
Run 3	8.50	8.50	8.50	8.50
Multi-agent average	8.90	9.00	8.67	8.89
Single-agent baseline	6.88	7.25	5.63	6.78

Table S3 : Multi-agent RaspAssistant performance and the single-agent baseline on Study 3.

Run	Workflow	Simulations	Analysis	Overall
Run 1	9.50	8.50	8.75	8.95
Run 2	9.00	8.31	8.75	8.68
Run 3	8.75	9.25	7.50	8.70
Multi-agent average	9.08	8.69	8.33	8.78
Single-agent baseline	9.38	8.63	5.00	8.20

Table S4 : Multi-agent RaspAssistant performance and the single-agent baseline on Study 4.

Agent	Type	Model	R	W	P	V	Other Tools
Supervisor	PEV	GPT-5.2	✗	✗	✗	✗	Agent delegation, fitting-loop entry
Setup	PEV	GPT-5.2	✓	✓	✗	✗	RASPA manual
Example	ReAct	GPT-5 mini	✓	✗	✗	✗	RASPA examples
File	ReAct	GPT-5 mini	✓	✓	✓	✗
Analysis	PEV	GPT-5.4	✓	✓	✗	✓	Output file parsing
Python	ReAct	GPT-5.2	✓	✓	✓	✗	Output file parsing

Table S5 : Agent architectures, configured model families, and capabilities. Parentheses indicate case-study configurations where they differ from workflow reproduction.

Dimension	Definition	Weight
Workflow correctness	Completion of the expected workflow steps in the appropriate order and using appropriate methodological choices.	40%
Simulation validity	Physical and computational validity of the simulations, including parameter selection and convergence.	40%
Analysis consistency	Internal support and calibration of the scientific interpretation, including uncertainty and model limitations.	20%

Table S6 : Workflow-reproduction evaluation dimensions and weights.

Dimension	Definition	Weight
Parameter correctness	Correctness and completeness of the extracted parameters relative to the reference parameter set.	60%
RASPA readiness	Fraction of generated force-field files that execute in RASPA without correction.	40%

Table S7 : Force-field extraction evaluation dimensions and weights.

ID	Full-credit requirement	Points
V1	Correct structures, species, temperatures, pressures, and compositions for the requested result families.	6
V2	Supplied force fields and charge convention are used effectively. Required specialized interactions are retained. Effective electrostatics are confirmed from output.	7
V3	Appropriate supercell replication and cutoff settings are used for the benchmark.	6
V4	Equilibration and production are credible, and convergence, uncertainty, or an equivalent stability check is demonstrated.	6
V5	Pressure, loading, heat, geometry, and normalization conventions are reconciled and comparisons are restricted to overlapping domains.	7
V6	Inputs, completed outputs, warnings, derived values, and report claims are traceable.	4

Table S8 : Simulation-validity criteria used for every workflow-reproduction benchmark.

ID	Full-credit requirement	Points
I1	Species, equations, signs, and reported values are internally correct.	5
I2	Claims follow the run’s usable outputs. A conclusion of “inconclusive” is accepted when supported.	5
I3	Small denominators, force-field sensitivity, conflicting models, and sampling limitations are recognized.	5
I4	Experimental comparison, pure fits, IAST, direct mixtures, breakthrough, and site evidence are weighted according to their reliability. Mechanistic claims have direct support.	5

Table S9 : Analysis-consistency criteria used for every workflow-reproduction benchmark.

ID	Expected workflow step	Points
TAM-W1	Inspect the structure, charges, experimental metadata, and distinct adsorption and column void fractions.	3
TAM-W2	Run pure CO 2 and N 2 GCMC at 298, 323, and 353 K over the experimental pressure range.	6
TAM-W3	Compare unit-matched pure simulations with experiment over the overlapping domains.	5
TAM-W4	Calculate selectivity and TSN over all four feeds, three temperatures, and the pressure range.	7
TAM-W5	Use IAST and representative direct mixture GCMC to assess mixture-model reliability.	4
TAM-W6	Compare breakthrough with the supplied pure-CO 2 /He and CO 2 /N 2 /He experiments.	6

Table S10 : Workflow criteria for TAMOF-1 CO 2 /N 2 separation 38 .

Figure S1 : TAMOF-1 diagnostic from Study 1, multi-agent Run 3, comparing IAST and direct-mixture GCMC selectivities for the benchmark based on Ref. 38 . Deviations from the parity line motivate treating IAST-derived selectivities and downstream breakthrough predictions as qualified rather than requiring numerical agreement.

ID	Expected workflow step	Points
ACE-W1	Inventory the four structures, material-specific force fields, and experimental states.	3
ACE-W2	Validate pure C 2 H 2 , CO 2 , and CH 4 isotherms at their native experimental temperatures.	6
ACE-W3	Validate adsorption heats using compatible loading bases and overlapping domains.	4
ACE-W4	Validate Fe-MOF-74 CO 2 and C 2 H 2 binding geometries at 298 K.	3
ACE-W5	Perform the 318 K pure-component study with host–guest and guest–guest evidence.	6
ACE-W6	Run equimolar mixture GCMC for both gas pairs at 298 K over the pressure range.	8

Table S11 : Workflow criteria for C 2 H 2 /CO 2 and C 2 H 2 /CH 4 separation 39 .

Figure S2 : Pure-component validation overview from Study 2, multi-agent Run 2, for the four MOF materials and supplied experimental reference data 39 . Material-dependent disagreement is not itself a scoring failure.

ID	Expected workflow step	Points
M74-W1	Audit all six materials and the experimental units and definitions.	3
M74-W2	Validate CO 2 at 298 K and H 2 at 77 K for all six metals.	6
M74-W3	Validate CO 2 and H 2 adsorption heats using the correct loading axes.	4
M74-W4	Validate CO 2 and H 2 binding geometries at 298 K.	4
M74-W5	Generate and audit 298 K pure CO 2 and H 2 data and fits for process calculations.	4
M74-W6	Compare IAST with representative direct mixture GCMC and judge reliability.	5

Table S12 : Workflow criteria for metal-center effects in M-MOF-74 40 .

Figure S3 : CO 2 isotherms at 298 K from Study 3, multi-agent Run 2, for the six M-MOF-74 materials. Black lines denote simulations and colored circles the experimental reference data 40 . Loadings are in mol kg -1 . The comparison illustrates both successful reproduction of broad adsorption trends and material-dependent deviations. Numerical agreement is not itself a scoring criterion.

ID	Expected workflow step	Points
ZEO-W1	Audit structures, cation and restart data, force fields, experimental states, units, and pressure coverage.	3
ZEO-W2	Calculate the C 3 –C 6 Rosenbluth prerequisites at 298 K and the C 6 values across the isobar temperatures.	4
ZEO-W3	Simulate and compare propane and propene isotherms for LTA5A and NaY at 298 K.	5
ZEO-W4	Simulate and compare hexane and 1-hexene isobars for LTA5A and NaY at the experimental pressures.	5
ZEO-W5	Run equimolar direct-mixture GCMC for all eight frameworks and four chain lengths at 298 K over the pressure range, including 100 kPa.	9
ZEO-W6	Calculate and audit selectivity versus pressure and at 100 kPa, including denominator-sensitive points.	5

Table S13 : Atomic workflow criteria for the olefin/paraffin zeolite benchmark.

Case	Source
EPM2 CO 2	Harris and Yung 51
CO 2 in zeolites	Garcia-Sanchez et al. 44
CO 2 in zeolites	Fang et al. 52
CO 2 , N 2 , O 2 , and Ar in zeolites	Vujic et al. 53
TraPPE-zeo	Bai et al. 54
Hydrocarbons and permanent gases in zeolites	Findley and Sholl 55

Table S14 : Scope of the force-field extraction benchmark.

Figure S4 : Post-fitting energetic assessment for C 2 H 2 in Fe-MOF-74 at 318 K. The experimental curve is from Bloch et al. 58 . The literature-force-field points were digitized from Ref. 39 . Experimental heats were unavailable to the fitting agent. RaspAssistant results use the final fitted force field. The zero-loading Widom prediction was monitored without an experimental heat reference or use in selection. Error bars show the reported simulation uncertainties and are smaller than the symbols where not visible.

Scientific Workflow Automation

Scientific workflow automation uses software, AI agents, and laboratory systems to plan, execute, validate, and reproduce research tasks.

See all 161 papers

SimMOF: AI agent for Automated MOF Simulations
Mar 31, 2026Jaewoong Lee, Taeun Bae, Jihan KimScientific Workflow Automation Materials Science
Department of chemical and Biomolecular Engineering, Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea
Mimosa Framework: Toward Evolving Multi-Agent Systems for Scientific Research
Mar 30, 2026Martin Legrand, Tao Jiang, Matthieu Feraud +5Autonomous Research Agents Multi-Agent Systems
Université Côte d’Azur, CNRS, ICN, Nice, France · Interdisciplinary Institute for Artificial Intelligence (3iA) Côte d’Azur, Sophia-Antipolis, France · Inria, Université Côte d’Azur, CNRS, I3S, France
VASP Agent: An Agentic Framework for Autonomous First-principles Calculations
Dec 22, 2025Zeyu Xia, Jinzhe Ma, Congjie Zheng +11AI Coding Agents Computational Materials Science
Shanghai Artificial Intelligence Laboratory, Shanghai, 200232, China. · Department of Computer Science and Technology, Tsinghua University, Beijing, 100084, China. · School of Physical Science and Technology, ShanghaiTech University, Shanghai, 201210, China. +6

Mar 31, 2026J/K move · Enter open · S save

SimMOF: AI agent for Automated MOF Simulations

Jaewoong Lee, Taeun Bae, Jihan Kim

Department of chemical and Biomolecular Engineering, Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea

Scientific Workflow Automation Materials Science Computational Materials Science Multi-Agent LLM Systems

Open paper AskPDF arXiv

Metal-organic frameworks (MOFs) offer a vast design space, and as such, computational simulations play a critical role in predicting their structural and physicochemical properties. However, MOF simulations remain difficult to access because reliable analysis require expert decisions for workflow construction, parameter selection, tool interoperability, and the preparation of computational ready structures. Here, we introduce SimMOF, a large language model based multi agent framework that automates end-to-end MOF simulation workflows from natural language queries. SimMOF translates user requests into dependency aware plans, generates runnable inputs, orchestrates multiple agents to execute simulations, and summarizes results with analysis aligned to the user query. Through representative case studies, we show that SimMOF enables adaptive and cognitively autonomous workflows that reflect the iterative and decision driven behavior of human researchers and as such provides a scalable foundation for data driven MOF research.

Abstract

Scientific Workflow Automation

Explore similar work

SimMOF: AI agent for Automated MOF Simulations

Mimosa Framework: Toward Evolving Multi-Agent Systems for Scientific Research

VASP Agent: An Agentic Framework for Autonomous First-principles Calculations