BET - Brunauer-Emmett-Teller

Brunauer-Emmett-Teller (BET) analysis is used to measure the specific surface area, pore volume, and pore size distribution of solid materials via gas adsorption. It provides key physical properties related to material reactivity and processing behavior.

Method Overview

Property	Value
Full Name	Brunauer-Emmett-Teller Surface Area Analysis
Purpose	Measures surface area, pore volume, and pore size distribution
Adsorptive Gas	Nitrogen (N2) at 77K
SharePoint Location	Analytical Data > BET
File Format	CSV (instrument export)

Key Metrics

Metric	Units	Description
AS(BET)	m²/g	Specific surface area via BET method
AS(BETMP)	m²/g	Multi-point BET surface area
Total Pore Volume	cm³/g	Total pore volume at specified p/p0
Average Pore Diameter	nm	Average pore diameter
C-Value	dimensionless	BET constant (adsorption energy indicator)

Pore Size Classification

Classification	Diameter Range
Micropores	< 2 nm
Mesopores	2-50 nm
Macropores	> 50 nm

Data Pipeline

SharePoint Sources

Type	Site	Folders	Files
BET Data	Analytical Data	BET, BET Data	.csv

File naming pattern: {sample_id} - BET Data Output.csv

Dagster Assets

The BET data pipeline consists of the following assets in apps/datasmart/src/datasmart/assets/analytical/bet.py:

bet_tables (sharepoint_multi_asset)
├── bet_params_incremental (backend.bet_params_incremental)
└── bet_raw (analytical.bet_raw)
          ↓
    bet (analytical.bet)
          ↓
    bet_simplified (analytical.bet_simplified)

Asset Descriptions

Asset	Schema	Description
bet_params_incremental	backend	Melted parameter data from CSV files
bet_raw	analytical	Raw isotherm data points
bet	analytical	Wide format, 1 row per sample with all parameters
bet_simplified	analytical	Key metrics only, C > 0 filtered, replicates averaged

Database Tables

analytical.bet_raw

Raw isotherm data points from BET measurements.

Column	Type	Description
sample_id	string	Normalized sample ID
original_sample_id	string	Raw sample ID from filename
No	integer	Data point number
p/p0	float	Relative pressure
p/Va(p0-p)	float	BET transform value
file_id	string	SharePoint file identifier

analytical.bet

Wide format table with all BET parameters.

Column	Type	Description
sample_id	string	Normalized sample ID
original_sample_id	string	Raw sample ID
analysis_date	datetime	Date of analysis
operator	string	Operator who performed analysis
as(bet)	float	BET surface area (m²/g)
as(betmp)	float	Multi-point BET surface area (m²/g)
c	float	BET C constant
total_pore_volume	float	Total pore volume (cm³/g)
average_pore_diameter	float	Average pore diameter (nm)
p/p0	float	Relative pressure for pore calculation
instrument	string	Instrument name
s/n	string	Serial number
analysis_port	string	Analysis port used
sample_mass	float	Sample mass (g)
adsorptive	string	Adsorptive gas type
adsorption_data_point_number	integer	Number of adsorption points
desorption_data_point_number	integer	Number of desorption points

analytical.bet_simplified

Simplified table with key metrics only.

Column	Type	Description
sample_id	string	Normalized sample ID
as(bet)	float	BET surface area (m²/g)
total_pore_volume	float	Total pore volume (cm³/g)
average_pore_diameter	float	Average pore diameter (nm)
p/p0	float	Relative pressure

File Format

CSV Structure

BET CSV files contain two sections:

1. Parameters section: Key-value pairs with parameter, value, and units columns
2. Raw data section: Isotherm data starting with "No" header

Example structure:

Instrument,BELSorp Mini II,
S/N,xxxxx,
Analysis date,2024-01-15,
Operator,John Doe,
Sample mass,0.1234,g
...
AS(BET),12.345,m²/g
Total pore volume at p/p0=0.99,0.0567,cm³/g
C,89.12,
...
No,p/p0,p/Va(p0-p)
1,0.05,0.123
2,0.10,0.234
...

C-Value Validation

The BET C constant indicates the validity of the measurement:

C Value	Interpretation
C > 0	Valid measurement
C < 0	Poor fit, non-ideal adsorption
Very large C	Potential issues with measurement

The bet_simplified asset automatically filters for measurements where C > 0.

Usage Examples

Query Simplified Data

from shared.db.sql import SQL

# Get simplified BET data (recommended for most analyses)
bet = SQL.read("""
    SELECT sample_id, "as(bet)", total_pore_volume, average_pore_diameter
    FROM analytical.bet_simplified
    WHERE sample_id LIKE 'P800%'
""")

Query Full BET Data

# Get full BET data with all parameters
full_bet = SQL.read("""
    SELECT sample_id, original_sample_id, analysis_date, operator,
           "as(bet)", c, total_pore_volume, average_pore_diameter,
           instrument, sample_mass
    FROM analytical.bet
    WHERE sample_id IS NOT NULL
    ORDER BY analysis_date DESC
""")

Query Isotherm Data

# Get raw isotherm data for plotting
isotherm = SQL.read("""
    SELECT "No", "p/p0", "p/Va(p0-p)"
    FROM analytical.bet_raw
    WHERE sample_id = 'P800-001-1-QBND-FM'
    ORDER BY "No"
""")

Plot BET Isotherm

import matplotlib.pyplot as plt

isotherm = SQL.read("""
    SELECT "p/p0", "p/Va(p0-p)"
    FROM analytical.bet_raw
    WHERE sample_id = 'P800-001-1-QBND-FM'
    ORDER BY "No"
""")

plt.figure(figsize=(10, 6))
plt.plot(isotherm["p/p0"], isotherm["p/Va(p0-p)"], 'o-')
plt.xlabel("Relative Pressure (p/p0)")
plt.ylabel("p/Va(p0-p)")
plt.title("BET Isotherm")
plt.grid(True, alpha=0.3)
plt.show()

Join with XRF Data

# Combine BET with XRF for material characterization
combined = SQL.read("""
    SELECT
        b.sample_id,
        b."as(bet)" as surface_area,
        b.total_pore_volume,
        b.average_pore_diameter,
        x.CaO, x.MgO, x.SiO2
    FROM analytical.bet_simplified b
    LEFT JOIN analytical.xrf_simplified x
        ON b.sample_id = x.sample_id
    WHERE b.sample_id IS NOT NULL
""")

Sample ID Processing

The pipeline uses the SampleID class to normalize sample identifiers:

from data_infrastructure.sample_id import SampleID

# Normalize sample IDs
df = SampleID.update_sample_id(
    df,
    is_material=True,
    project_team_map=project_team_map,
    crosswalk=sample_id_crosswalk,
)

Related Assets

Asset	Description
sample_id_crosswalk	Maps messy sample IDs to canonical format
project_team_map	Maps old process area codes to new project team codes

Troubleshooting

Missing Samples

If a sample is missing from BET tables:

1. Check if the source CSV exists in SharePoint under BET/BET Data/
2. Verify the file naming follows {sample_id} - BET Data Output.csv
3. Check backend.bet_params_incremental for raw data

Invalid C Values

If samples are missing from bet_simplified:

1. Check the c value in analytical.bet - must be > 0
2. Negative or zero C values indicate measurement issues
3. Contact the analytical team for re-measurement if needed

# Find samples filtered out due to C value
SQL.read("""
    SELECT sample_id, "as(bet)", c
    FROM analytical.bet
    WHERE c <= 0
""")

Sample ID Normalization Issues

If sample IDs aren't matching:

1. Check original_sample_id column for the raw value
2. Review sample_id_crosswalk for applicable mappings
3. Update crosswalk if a new mapping is needed