Surface Analysis Descriptors

This page describes the Surfacia descriptors that come directly from molecular surface electronic structure and from multi-scale quantitative surface analysis.

Overview

Surface descriptors are central to Surfacia because many chemically important processes are expressed at the molecular surface:

  • intermolecular recognition

  • steric accessibility

  • electrostatic complementarity

  • local electron donation and acceptance

Core Surface Properties

Surfacia uses three main surface-electronic quantities.

ESP (Electrostatic Potential)

  • Describes charge distribution on the molecular surface

  • ESP_min usually marks the most electron-rich region

  • ESP_max usually marks the most electron-deficient region

ALIE (Average Local Ionization Energy)

  • Reports local resistance to electron removal

  • Lower values usually indicate easier electron donation

LEAE (Local Electron Attachment Energy)

  • Reports local tendency to accept electrons

  • Useful for identifying electrophilic character and local acceptor strength

Common Statistics

For a given property, Surfacia frequently reports:

  • *_min: most extreme low-value region

  • *_max: most extreme high-value region

  • *_mean or *_average: overall character

  • *_delta: heterogeneity across sites

  • *_variance: statistical spread across the surface

Three MQSA Modes

Mode 1: Element-Specific

Use this mode when one element is already known to be chemically central.

  • generates 13 descriptors

  • aggregates surface properties over atoms of a chosen element

  • useful for element-aware hypothesis testing

Mode 2: Fragment-Specific

Use this mode when you already know the important fragment, catalytic core, or pharmacophore.

  • generates 18 descriptors

  • uses the naming pattern Fragment_[Property]_[Statistic]

  • captures how a fixed local motif is perturbed by the surrounding structure

Mode 3: LOFFI Automated

Use this mode when you want broad exploratory analysis without imposing a mechanism first.

  • generates 32 descriptors

  • combines atom-level and functional-group-level summaries

  • is best suited to diverse molecular datasets

How to Read the Names

Atom_ descriptors

  • global atom-level summaries over the full molecular surface

  • example: Atom_ALIE_mean

Fun_ descriptors

  • statistics over automatically detected functional groups

  • example: Fun_ESP_delta

Fragment_ descriptors

  • statistics for a user-defined fragment

  • example: Fragment_ESP_mean

Element-centered descriptors

  • statistics aggregated over a selected element in Mode 1

  • example: element-specific area, min, max, mean, and delta features

What the Modes Are Good For

Each mode answers a slightly different question:

  • Mode 1: "Is this particular element driving the chemistry?"

  • Mode 2: "How do substituents perturb a shared reactive fragment?"

  • Mode 3: "What local and global surface features matter when I do not want to assume a mechanism first?"

Why These Descriptors Matter

In practical modeling, surface descriptors often provide the most chemically interpretable bridge between:

  • quantum calculations

  • compact machine-learning models

  • SHAP explanations

  • human-readable design insight

They are especially useful when you want to explain not only whether a molecule performs well, but also which regions and which physicochemical patterns are likely responsible.