Modular Electrochromic Polyester (MEP) Color Targeting

Python tools for predicting and targeting colors obtained from blends of modular electrochromic polyesters (MEPs) using the CIE 1976 L* a* b* (CIELAB) color space.

Core capabilities:

• Convert experimental transmission spectra of MEP devices into L* a* b* coordinates.
• Predict colors from mixtures of two MEPs using linear interpolation of spectra in CIELAB space.
• Predict colors from mixtures of three MEPs using barycentric coordinates in the a*–b* plane.
• Compute ΔE color differences between predicted mixtures and a user-defined target color.
• Suggest “best starting ratios” of MEPs for experimental color targeting.

---

Background

MEPs are electrochromic polyesters built from aromatic ester macromonomers (e.g. terephthalate, naphthalate, thiophenate) linked by flexible aliphatic segments (e.g. adipate, hexamethylene). They combine:

• Bright, high-contrast colors in the reduced state (PHAT = magenta, PHAN = green, PHATh = cyan, etc.).
• High neutral-state transparency.
• Good processability and chemical recyclability via ester hydrolysis.

In the reduced state, blends of MEPs behave approximately linearly in CIELAB space: the color of a blend lies along (2 MEPs) or inside (3 MEPs) the convex hull spanned by the individual MEP colors. This makes them ideal for model-based color targeting instead of brute-force experimental screening.

All color calculations are performed in CIE 1976 L* a* b*:

• L* – lightness (0 = black, 100 = white)
• a* – green (−) to red (+)
• b* – blue (−) to yellow (+)

Color differences are quantified using the CIE76 metric:

Delta E = sqrt{(L_2* - L_1*)^2 + (a_2* - a_1*)^2 + (b_2* - b_1*)^2}

---

Methods Overview

2-MEP Targeting (Linear Interpolation)

For two MEPs with spectra A_1(λ) and A_2(λ), the code constructs a family of theoretical spectra:

A_th(λ, k_i) = α_i A1(λ) + (1-α_i) A2(λ)

with α_i in [0,1] sampled at user-defined resolution (e.g. 5% or 10% steps).

Each spectrum is:

1. Converted to Lab* (given an illuminant and observer).
2. Compared to the target Lab* color via ΔE.
3. Ranked to return the MEP mixing ratio that minimizes ΔE.

3-MEP Targeting (Barycentric Coordinates in a*–b*)

For three MEPs (for example PHAT, PHAN, PHATh) with reduced-state coordinates A = (a_A*, b_A*), B = (a_B*, b_B*), C = (a_C*, b_C*) and a target color P = (a_P*, b_P*), the code solves for barycentric weights lambda_A, lambda_B, lambda_C such that

P ≈ lambda_A A + lambda_B B + lambda_C C

with

lambda_A + lambda_B + lambda_C = 1 and lambda_i >= 0 for all i.

The weights directly correspond to the blend ratio of each MEP. In the current implementation, the targeting is carried out in the a*–b* plane (hue). The same framework naturally extends to four MEPs in full 3D L* a* b* space (tetrahedral barycentric coordinates) if one wishes to incorporate the L* axis in the targeting.

---

Repository Structure

MEP/ │ ├─ 2MEP targeting/ │ ├─ PHAN.csv │ ├─ PHAT.csv │ ├─ PHATh.csv │ ├─ color_coor_ref.csv │ ├─ photopic_ref.csv │ └─ theo.py │ ├─ 3MEP targeting/ │ ├─ bary-targeting.py │ ├─ color_coor_ref.csv │ ├─ photopic_ref.csv │ └─ real.csv │ ├─ README.md └─ LICENSE