# Complex Metal Oxide Optical Spectroscopy from Automated Inkjet Synthesis

## Facts

- Provider: Google Research
- License: CC-BY-4.0 (https://storage.googleapis.com/gresearch/metal-oxide-spectroscopy/LICENSE)
- Format: CSV (in sampled files)
- Stored at: Google Cloud Storage, gs://gresearch/metal-oxide-spectroscopy (us)
- Access: public, readable without an account
- Category: Chemistry & Materials
- Source: https://storage.googleapis.com/gresearch/metal-oxide-spectroscopy/README.txt
- Page: https://scigantic.com/archive/1cf74a91-28f3-556e-b628-87c06fb8fd51
- Keywords: Chemistry & Materials, spectroscopy, CSV, Google Cloud Storage

## About

Google Research dataset of optical transmission spectra measured on inkjet-printed complex metal oxide composition libraries, for high-throughput materials discovery. Per the publisher README, it covers 6 print sessions, 173 plates, 131 quaternary oxide systems, 6,918,024 individual composition samples and 376,752 distinct compositions drawn from the metals Ce, Co, Cu, Fe, In, Mg, Ni, Sn, Ta and Y, with optical transmission measured at 9 wavelengths from 375 to 850 nm and tenfold replicate patches within each plate. Four data types are given as one CSV per plate: ten_replicas (per-patch spectra with metal fractions and plate coordinates), aggregated_replicas (replicates averaged), mixture (log probability that a composition is explained by at most 3 binary signals) and phase_fits (fitted phase diagrams with fitted absorption spectra per phase point), plus metadata.csv listing each plate's anneal temperature, print session, ink channels and design, and microscope images. Measured 2026-11-13: folders aggregated_replicas, images (PNG per plate), mixture, phase_fits and ten_replicas, with per-plate ten_replicas CSVs around 10.8 MB each. Read the CSVs with pandas. The README notes that consistency between print sessions is weaker than within a plate. Licensed CC BY 4.0 per the in-prefix LICENSE; cite the Google Research publication when it appears.

## Load the data

```python
import os
import pandas as pd
df = pd.read_csv(os.path.join(os.environ['SCIGANTIC_MOUNT_PATH'], 'aggregated_replicas/45706.csv'), nrows=1000)
df.head()
```

Run in a Scigantic notebook on 2026-11-14 against `aggregated_replicas/45706.csv`: CSV/TSV: columns ['Fe', 'Mg', 'Ta', 'signal_0', 'signal_1', 'signal_2', 'signal_3', 'signal_4', 'signal_5', 'signal_6', 'signal_7', 'signal_8'], dtypes ['float64', 'float64', 'float64', 'float64', 'float64', 'float64', 'float64', 'float64', 'float64', 'float64', 'float64', 'float64'], first 1000 row.

From any machine:

```python
import gcsfs
fs = gcsfs.GCSFileSystem(token='anon')
fs.ls('gresearch/metal-oxide-spectroscopy/', detail=False)[:10]
```

Lists the files with anonymous access. Runs on any machine, with no cloud account.

## Data structure

File types in sampled files: CSV

Example file: `metal-oxide-spectroscopy/aggregated_replicas/45706.csv`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/1cf74a91-28f3-556e-b628-87c06fb8fd51
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/1cf74a91-28f3-556e-b628-87c06fb8fd51
- Over MCP: `get_schema_card` with id `1cf74a91-28f3-556e-b628-87c06fb8fd51` at https://api.scigantic.com/mcp

## Related topics

- [CSV datasets](https://scigantic.com/topics/csv)
- [Chemistry & Materials datasets](https://scigantic.com/topics/chemistry-and-materials)
- [Spectroscopy datasets](https://scigantic.com/topics/spectroscopy)
