Computational UV-vis Spectroscopy of Food Dyes

Computational chemistry lesson. Students run TDDFT calculations in Turbomole/TmoleX on an HPC cluster to predict the UV-vis spectra and colors of three candidate molecules (A, B, C) and identify which one is the Red-40 dye, by comparing the computed spectra/colors with the dye's known experimental absorption (maximum near 504 nm). The cyberinfrastructure skills are remote HPC access via Open OnDemand and setting up/running a quantum-chemical calculation through a GUI. There is no wet-lab component in this lesson.

This README is the technical setup document: lesson information, instructor setup, software installation/licensing, and implementation notes. The student-facing lesson is student.md; the answer key, worked solution, and in-context teaching tips are in instructor.md.

Lesson Information

The Lesson Information below should be kept in sync with the student lesson.

Lesson Learning Outcomes

At the end of this lesson, students will be able to:

1. Import a molecular structure and set up a quantum-chemical calculation in TmoleX.
2. Choose appropriate calculation settings: basis set, molecular charge, density functional, and implicit solvation.
3. Run TDDFT calculations on an HPC cluster through a job queue.
4. Compute a UV-vis spectrum and predict a molecule's color from it.
5. Compare computed spectra and colors with reference data to identify a compound.
6. Relate the size of a π-conjugated system to its absorption wavelength.

Cyberinfrastructure Prerequisites

• Ability to log in to an HPC cluster through an Open OnDemand web portal.
• Comfort launching a remote desktop session and using a graphical application within it.
• Basic file handling on a remote system (locating and copying input files).
• No programming is required; the lesson is a guided GUI workflow.

Content Prerequisites

• UV-vis absorption and electronic excitation as the origin of molecular color (introductory physical chemistry).
• Basic familiarity with molecular structure and π-conjugation.
• DFT/TDDFT are used as tools and explained conceptually; no prior quantum-chemistry coursework is assumed.

Resources

• Introduction to DFT (Grotjahn Lab)

Lesson Version & Intended Modality

A single student version is provided: student.md, written as a Markdown lab guide with callouts and screenshots. It is designed to be used in lab with an instructor present (guided inquiry); it can also support asynchronous review. There is no executable code; "running" the lesson means viewing/sharing the Markdown.

Note: this departs from the cookiecutter default of three notebook versions plus separate instructor-key.ipynb / instructor-notes.ipynb. Per project decision, the lesson is Markdown with one student file, and all instructor material lives in this README.

Instructor Setup

The computational part assumes a Turbomole/TmoleX installation reachable through an HPC Open OnDemand portal. The student lesson is written against SCU's WAVE cluster but uses portable paths so it can be adapted elsewhere.

1. Accounts & network. Ensure each student has a cluster account and portal access, and knows the network requirement (e.g., eduroam).

2. Resource profile. Identify the Open OnDemand interactive "Compute GUI" profile that exposes the TmoleX launcher, and tell students to select it (a different profile may hide the launcher).

3. TmoleX availability. Confirm TmoleX starts from Applications → Other → TMOLEX. Identify the fallback launch path that students can type in the terminal /YOUR/INSTALLATION/PATH/TMOLEX/TmoleX/TmoleX25. Update this for your site.

4. Input files. Distribute the provided material/ files (A.xyz, B.xyz, C.xyz, remote-configuration.xml). The student lesson instructs copying them to /tmp, which exists on essentially any cluster:

   cp material/A/A.xyz material/B/B.xyz material/C/C.xyz material/remote-configuration.xml /tmp/

   You may instead pre-stage them in a shared course directory and adjust the lesson path accordingly.

5. material/remote-configuration.xml. This shipped file is a template for a SLURM cluster (TmoleX "Remote Systems" import). It is written with UPPERCASE placeholders and inline comments; before distributing it to students, open it and replace each placeholder with your site's values:

   • MachineGroupName: any label for the course or machine group.
   • Path: a shared scratch directory writable by all course users.
   • Turbodir: your TURBOMOLE installation directory (the value of $TURBODIR).
   • TurbomoleVersion: the installed TURBOMOLE version (for example, 8.0).
   • In the SLURM header: --partition=YOUR_PARTITION, the optional --account line (uncomment if your site requires an allocation), and the --mail-user domain.

   MachineName is 127.0.0.1, which works when jobs are submitted from inside the cluster (e.g. the Open OnDemand desktop); use your login-node hostname otherwise. The default request is 48 --cpus-per-task on one node, matching the lesson. The file contains no passwords, and each student enters their own credentials in TmoleX after importing it.

Obtaining a free educational Turbomole/TmoleX license

Turbomole offers a free educational license (TURBOMOLE 8.0 + the integrated TmoleX GUI) for teaching, coursework, and training:

1. Open the product page: https://store.turbomole.org/product/turbomole-8-0-educational/.
2. Download the educational license agreement (linked on that page), sign it electronically, and email the signed agreement to order@turbomole.org together with proof of student/educational status (enrollment certificate or student ID).
3. The license is free of charge and covers TURBOMOLE 8.0 for Linux, Windows, and macOS with the integrated TmoleX GUI, plus the standalone Linux build, with unlimited CPU/GPU cores.

Important

The educational license may not be used for scientific publishing (no publications, preprints, or conference proceedings). For published research, obtain a regular Turbomole license.

Installing Turbomole (command-line build)

Full instructions: https://manual.turbomole.org/v8.0/installation-of-turbomole/install-turbomole-command-line-version/. Summary for a shared multi-user install on a cluster or lab server:

1. Do not install or run Turbomole as root, because file ownership and permissions will be wrong. Unpack once as a normal user (ideally one whose group matches all course users) onto shared/network storage available on every compute node:

   gunzip turbomole_80.tar.gz
   tar -xvf turbomole_80.tar
   # or, for the self-extracting installer:  ./turbolinux80.bin

2. Per-user environment. Point TURBODIR at the unpacked directory and source the setup file (add these to each user's .bashrc/.profile):

   export TURBODIR=/shared/software/TURBOMOLE
   source $TURBODIR/Config_turbo_env          # use Config_turbo_env.tcsh for csh/tcsh

3. Verify the install:

   cd $TURBODIR/TURBOTEST
   TTEST

Student access options

Choose whichever fits your site; this TDDFT lesson is the more demanding of the two modules (it requests ~48 threads), so a cluster/queue is recommended over individual laptops.

• Option A: Browser access via Open OnDemand (recommended for this lesson). Install Turbomole/TmoleX on an HPC cluster that runs Open OnDemand and expose an interactive "Compute GUI" desktop app. Students open the portal in a web browser, launch the remote desktop, and run TmoleX there, with no local installation, and the heavy job runs on cluster nodes through the queue. This is the workflow the lesson screenshots show. Ensure the chosen resource profile exposes the TmoleX launcher (see step 2 above).
• Option B: Local TmoleX install on university computers. Install the educational TmoleX package (Linux/Windows/macOS) on managed lab/university machines from the license download, and have students run TmoleX locally. This is simplest for the companion IR lesson (small, local jobs); for this TDDFT lesson, local machines must be powerful enough for the larger calculation, otherwise prefer Option A.

Calculation settings (must match across all three structures)

• Basis set: def2-TZVPD (not def2-TZVPP).
• Molecular charge: −2 (Generate MOs after setting).
• Functional: PBE.
• Solvation: COSMO, ε = 80.4, refractive index = 1.33 (water).
• Property: UV/Vis and CD (vertical excitations), 40 excitations, All IRREPs.
• Execution: queued on the remote node (48 threads), not local.

Answer Key & Worked Solution

Important

The answer key, the formative-question answers, the worked solution (computed spectra, colors, and interpretation), and the in-context teaching tips now live in instructor.md, a single self-contained instructor document that mirrors the student lesson. Keep instructor.md out of any student-facing branch. This README holds the technical setup only.

Implementation Notes (for adopting instructors)

• Lesson timing: about 1.5 to 2 hours of computational work during the lab session.
• Group size / facilitator ratio: piloted with 1 instructor and 1 TA for 22 students in a single lab session.