Gathered here are links to data sets that can be used to illustrate topics in analytical chemistry. Each data set is available as a .csv spreadsheet.
Visible Spectra of Multicomponent Mixtures
Filename
CuCoCrNi.csv (download)
File Structure
Visible spectra recorded at 642 wavelengths (380.5 nm – 899.5 nm) for 80 solutions that contain 1 – 4 of the following transition metal ions: Cu2+, Co2+, Cr3+, or Ni2+ in 0.10 M HNO3. The spectra were recorded using a Vernier SpectroVis Plus spectrometer using LoggerPro and exported as a .csv file. Each row is a separate sample; the columns contain the following types of data:
column a: id for sample
column b: list of analytes in the sample
column c: number of analytes in the sample (labeled as dimension)
column d: concentration of Cu2+ (prepared from single stock solution of 0.0500 M Cu2+)
column e: concentration of Co2+ (prepared from single stock solution of 0.100 M Co2+)
column f: concentration of Cr3+ (prepared from single stock solution of 0.0375 M Cr2+)
column g: concentration of Ni2+ (prepared from single stock solution of 0.130 M Ni2+)
columns h – xr: absorbance at the stated wavelength
Suggested Uses
analysis for single analytes using external standards calibrations
multicomponent analysis of binary, ternary, and quaternary mixtures
choosing wavelengths for Beer’s law analysis
principal component analysis
cluster analysis
multiple linear regression
Resources That Use This Data Set
Using R to Introduce Students to Principal Component Analysis, Cluster Analysis, and Multiple Linear Regression (talk delivered at Pittcon 2018)
UV Spectra for Benzene
Filenames
File Structures
Each file contains a set of UV spectra (240– 290 nm) for a solution of benzene dissolved in cyclohexane recorded using a double-beam, scanning spectrophotometer (Perkin-Elmer Lambda 4B Spectrophotometer). Additional details on each file are provided here:
benzene_response.csv: All spectra were recorded using a scan rate of 60 nm/s and a slit width of 1.0 nm, and variable response times. The instrument gives the response times as a number from 1–7, which correspond to response times of 1 = 0.1 s, 2 = 0.2 s, 3 = 0.5 s, 4 = 1.0 s, 5 = 2.0 s, 6 = 5.0 s, and 7 = 10.0 s
benzene_scanrate.csv: All spectra were recorded using a response time of 05. s and a slit width of 1.0 mm, and variable scan rates of 5 nm/s, 10 nm/s, 20 nm/s, 60 nm/s, 120 nm/s, 200 nm/s, 300 nm/s, 750 nm/s, and 1500 nm/s
benzene_slitwidth.csv: All spectra were recorded using a response time of 0.1 s and a scan rate of 20 nm/s, and variable slit widths of 0.25 nm, 1.0 nm, 2.0 nm, and 4.0 nm.
Suggested Uses
explore effect of scan rate, slit width, and response time on the resolution of the peaks in benzene's spectrum
explore effect of scan rate, slit width, and response time on signal-to-noise ratio
explore effect of scan rate, slit width, and response time shape of spectral peaks, including height, width, and position
explore effect of scan rate, slit width, and response time on analysis time
discuss trade-offs between conditions that favor a more rapid analysis time, signal-to-noise ratio, and accuracy of peak shape and position
M & M Data
FIlename
MMdata.xlsx (download)
File Structure
Data recorded for 30 1.69-oz bags of plain M & Ms. The variables are:
the store where the bag was purchased (three stores, 10 sample each; stores were CVS, Kroger, and Target)
number of blue M & Ms in a bag
number of brown M & Ms in a bag
number of green M & Ms in a bag
number of orange M & Ms in a bag
number of red M & Ms in a bag
number of yellow M & Ms in a bag
number of yellow M & Ms in the first five samples drawn from a bag
net weight (in grams)
Suggested Uses
visualizing data using, for example, boxplots, bar plots, dot plots, and histograms
summarizing data using means, medians, standard deviations, and variances
modeling data using binomial, poisson, and normal distributions
examining confidence intervals and significance testing
Resources That Use This Data Set