Before You Start

Have the intended chemical form ready: structure, stereochemistry, charge or salt state, mixture composition, material role, and application matrix.
Collect measured flash point, boiling point, logP, pH, decomposition, environmental, and performance data with units and method references if you plan to override estimates.
Decide whether the task is single-molecule triage, mixture screening, regulatory review, decomposition assessment, field deployment, or safer-alternative design.
Use a modern browser with JavaScript enabled. Structure, prediction, analogue, batch, and export actions depend on the application services being available.

Step-By-Step Tutorial

Open ChemrytSC and draw the molecule in JSME, paste a SMILES, SELFIES, InChI, or name, or click `Molecule Demo` to load the example.
Click `Load Structure`, then confirm the normalized identity and 2D/3D structure. Enter the material name, application context, and optional measured flash point, boiling point, logP, and pH values.
For a mixture, add each component and enter its percentage plus an optional SMILES or CAS identifier. Confirm that the composition and component identities represent the intended formulation.
Click `Analyze Molecule` to generate the rule-based specialty-chemical review. Use `Run Prediction` for model endpoint predictions.
Review `Structure` and `SC Core` first for identity, descriptors, thermal profile, functional groups, formulation estimates, GHS guidance, and jurisdiction-focused regulatory screening.
Use `Decomposition`, `Environmental`, and `Field` to assess thermal breakdown, compartment fate, aquatic or bioaccumulation cues, deployment conditions, compatibility, and what-if variants.
Use `Models`, `Analogue Finder`, and `SELFIES Ensemble` to examine endpoint confidence, compare functional alternatives, and explore structure variants.
Export Markdown, the regulatory summary, or the field brief when needed, and record the exact structure, inputs, assumptions, evidence status, and validation actions with the project record.

Workflow A — Analyze a Single Specialty Chemical

Use this workflow for an additive, plasticizer, stabilizer, coating ingredient, solvent, intermediate, or performance chemical.

Draw the exact structure in JSME or paste the SMILES, SELFIES, InChI, or recognized chemical name in the molecule field.
Select `Load Structure`. Check that the editor, identity readout, canonical representation, formula, charge, fragments, and 3D view describe the intended material.
Enter the application context and add measured flash point, boiling point, logP, or pH only when reliable values are available. Leave `Auto` fields empty to retain calculated estimates.
Select `Analyze Molecule`. Wait for the result workspace to open, then resolve any parsing, service, or identity warning before continuing.
Read Structure identity and mapped features first, SC Core second, and the specialized tabs only after the submitted form has been confirmed.
Select `Run Prediction` when model endpoints are needed. Open model rationale or confidence details before using an endpoint in a decision.

Workflow B — Screen a Multi-Component Formulation

Use this workflow when formulation behavior or combined component alerts matter.

Load the principal component, then use `Add Component` for every additional substance that can affect performance, hazard, compatibility, or compliance.
Enter a component name, percentage, and optional SMILES or CAS. Use one composition basis—such as wt%—throughout and confirm the total.
Analyze the formulation and review component-specific identity or alert findings before interpreting combined guidance.
Use SC Core formulation estimates and HSP matching to explore matrix compatibility; select the intended target matrix and process temperature.
Use Field compatibility, suggested solvent systems, polymer matrices, additive watchouts, stability outlook, and deployment readiness to create a lab-test plan.
Do not assign a mixture GHS classification solely from molecular screening; apply validated mixture rules and measured data where required.

Workflow C — Find and Compare Safer Alternatives

Use this workflow after the parent material has been fully analyzed.

Open `Analogue Finder` and generate functional candidates. Review similarity, changed structural features, predicted property movement, and alert differences.
Use the side-by-side comparison to ensure the candidate retains the required specialty function rather than optimizing only hazard signals.
Open `SELFIES Ensemble` to generate additional valid string variants, then inspect every decoded structure and identity.
Use Environmental bioisostere suggestions or Field what-if variants to test targeted replacements and review the changed results.
Shortlist candidates only after checking synthesis or sourcing feasibility, application performance, stability, hazard, environmental fate, and jurisdictional status.

Complete Module Guide

Module	What to review	How to use the result
Structure	Canonical identity, SELFIES, descriptors, 2D/3D views, thermal profile, regioisomer or component triage, and RDKit structure map.	Verify the submitted chemical form and locate the structural basis of later alerts.
SC Core	Quick risk profile, functional groups, application tags, formulation properties, Hansen parameters, volatility, safer alternatives, GHS, handling, and global inventory overlays.	Build the primary formulation, safety, and compliance screening summary.
Decomposition	TDA snapshot, atmosphere and matrix settings, mass-loss threshold, kinetic method, weight-loss steps, vapor evolution, TGA/DTG simulation, fragments, and mitigation suggestions.	Prioritize thermal-analysis experiments and decomposition-product identification.
Environmental	Fate by compartment, aquatic hazard, metabolite hypotheses, BCF/BAF cues, restriction watchdog, PNEC/DNEL screening, replacements, and GHS overview.	Define environmental testing, exposure assessment, and regulatory follow-up.
Field	Batch variant scoring, structural what-if tools, analogue matrix, performance radar, safety brief, hydrolysis or ambient degradation, compatibility, deployment readiness, and field actions.	Translate molecule-level evidence into the intended application and operating conditions.
Models	Physicochemical, release-relevant, ADME, environmental, and toxicology endpoint predictions plus method and confidence information.	Use endpoints as ranked screening evidence and investigate out-of-domain or unavailable models.
Analogue Finder	Functional alternatives, similarity, changed motifs, comparison metrics, and copyable analogue SMILES.	Select candidates for deeper design, sourcing, or experimental work.
SELFIES Ensemble	SELFIES encoding and generated structure variants with session history.	Explore valid structural neighborhoods; verify every decoded molecule before use.

Worked Example

Input

Load a representative specialty additive, assign its intended polymer or coating application, and retain calculated property fields unless measured values are traceable.

Primary review

Confirm identity in Structure, then record SC Core functional groups, HSP match, volatility, GHS cues, jurisdiction settings, and handling guidance.

Deep review

Select the intended decomposition matrix and atmosphere, inspect environmental fate and field compatibility, and document the tests needed to resolve warnings.

Decision

Compare at least one analogue, state which function must be preserved, and export a summary that separates measured, predicted, rule-based, and unresolved evidence.

Troubleshooting

Problem	Likely cause	What to do
Structure does not load	The identifier is unsupported, ambiguous, malformed, or represents multiple fragments unexpectedly.	Use a valid explicit SMILES or redraw the structure; verify charges, ring closures, stereochemistry, dots, and counterions.
Analyze or prediction fails	A chemistry service, model service, or Python runtime may be unavailable, or the structure may not parse.	Preserve the input, read the visible error, correct structure issues, retry once, and leave failed endpoints unresolved if the service remains unavailable.
2D and 3D forms look different	Normalization, aromaticity, protonation, salt handling, or 3D embedding can change the display.	Compare the canonical identity and submitted form; do not proceed until the scientifically intended representation is confirmed.
Results change after entering a measured value	The measured-property override is intentionally used by downstream rules.	Verify units and method, remove an untrusted override, then rerun analysis and document the change.
Mixture output is misleading	Components, percentages, identifiers, or composition basis may be incomplete or inconsistent.	Correct the component list and total, analyze components individually, and apply validated mixture-classification rules.
Export is empty or outdated	Analysis was not completed after the last structure or context change.	Rerun analysis or prediction, confirm the active tab shows current results, and generate the export again.

Tutorial Notes

Use measured-property overrides only when their method, units, temperature, and source are known; an override changes downstream interpretation.
For mixtures, analyze every relevant component and check that percentages represent the same basis before interpreting formulation or hazard results.
A structural alert or restriction watchdog match is a screening signal, not proof that a substance is legally listed or prohibited in a jurisdiction.
Review model status, applicability, and rationale. Keep unavailable or degraded endpoints visibly unresolved instead of replacing them with assumptions.
Generated analogues and SELFIES variants require identity, synthesizability, performance, safety, and regulatory validation before selection.

Good Practice

ChemrytSC provides structure-led screening and decision support. Confirm composition, performance, hazards, environmental fate, legal status, and handling requirements using validated measurements, current jurisdictional sources, SDS review, and qualified expert assessment.