A chromatogram is a graph that plots detector response on the vertical axis against time on the horizontal axis, with each peak representing a different compound in your sample. Reading one comes down to three questions: what’s in the sample (identified by where a peak appears), how much is there (determined by peak size), and whether the separation worked well (judged by peak shape and spacing). Once you understand these basics, even a complex chromatogram with dozens of peaks becomes interpretable.
The Two Axes
The horizontal axis (x-axis) shows retention time, measured in minutes. Retention time is how long a compound took to travel from the injection point, through the column, and to the detector. Compounds that interact weakly with the column material pass through quickly and appear on the left side of the chromatogram. Compounds that interact strongly with the column take longer and appear further to the right.
The vertical axis (y-axis) shows detector response, which reflects how much of a substance the detector is sensing at any given moment. The units depend on the detector type: absorbance units for UV detectors, millivolts for some others. For practical purposes, a taller signal means more substance is hitting the detector at that instant.
What Peaks Tell You
Each peak on a chromatogram represents a compound passing through the detector. The tip of the peak, called the apex, marks the moment when the highest concentration of that compound reaches the detector. The retention time at the apex is the number you use to identify what the compound is.
Identification works by comparison. If you run a known standard (say, pure caffeine) under the same conditions and it produces a peak at 4.2 minutes, then a peak at 4.2 minutes in your unknown sample is likely caffeine. This only works when the column, temperature, flow rate, and mobile phase composition are identical between runs. Change any of those variables and retention times shift.
The baseline is the flat line running along the bottom of the chromatogram when no compounds are being detected. Peaks rise above this baseline and, ideally, return cleanly to it before the next peak begins. When two peaks overlap and don’t return to baseline between them, it’s called co-elution, and it makes both identification and quantification less reliable.
Measuring How Much Is There
Quantification relies on peak area, not peak height. Peak area is the total space under the curve of a peak, calculated automatically by the instrument’s software using integration. The reason area is preferred: two peaks can have the same concentration but very different heights if one is broad and the other is narrow. Area accounts for both the height and width of the peak, giving a more accurate reflection of the total amount of compound present.
Peak height was standard practice decades ago when labs relied on strip chart recorders and analysts literally cut out paper peaks and weighed them. Modern computerized integrators calculate area under the curve precisely, making height-based measurements largely obsolete. The one exception is when baseline noise is significant. In noisy chromatograms, peak height can sometimes be a more reliable measurement than area, particularly when determining whether a trace compound is even detectable.
To convert peak area into an actual concentration, you need a calibration curve. This means running several standards of known concentration, plotting their peak areas against concentration, and fitting a line. Your unknown sample’s peak area then maps onto that line to give you a concentration value.
Internal Standards
An internal standard is a known amount of a compound, different from your analyte, that you add to every sample before injection. It appears as its own peak on the chromatogram. Instead of using raw peak areas, you calculate the ratio of your analyte’s peak area to the internal standard’s peak area. This ratio corrects for small variations in injection volume or instrument response between runs, improving accuracy considerably.
Judging Peak Shape
A perfect peak is symmetrical and Gaussian, shaped like a bell curve. Real peaks rarely look perfect, and the way they deviate tells you something about what’s going on in the system.
Tailing occurs when the back half of the peak stretches out longer than the front half, creating a shape like a shark fin. The most common causes are column overload (too much sample injected), a chemically heterogeneous column surface, or extra-column effects like dead volume in fittings and connections. In gas chromatography specifically, tailing can appear when the injection port temperature isn’t high enough to vaporize the sample quickly.
Fronting is the opposite: the leading edge of the peak stretches forward. This typically happens when the column is overloaded with a particular compound, or when the sample solvent is stronger than the mobile phase.
Peak asymmetry is measured by comparing the width of the back half of the peak to the front half, taken at 10% of the peak height. A perfectly symmetrical peak has an asymmetry factor of 1.0. Values above 1.0 indicate tailing; values below 1.0 indicate fronting. Most labs consider asymmetry factors between 0.8 and 1.5 acceptable for routine work.
Resolution Between Peaks
Two peaks that are close together need enough separation to be measured independently. Resolution describes how well two adjacent peaks are separated. Visually, good resolution means the signal returns close to baseline between two peaks. Poor resolution means the peaks merge into each other, making it hard to draw accurate boundaries for integration.
Three factors control resolution: column efficiency (how narrow the peaks are), selectivity (how different the retention times are), and the retention factor. You can often improve resolution by using a longer column, changing the mobile phase composition, or adjusting the temperature.
The retention factor, commonly written as k, tells you how long a compound spends interacting with the column material relative to simply flowing through. It’s calculated as the retention time of your compound minus the time it takes an unretained compound to pass through, divided by that unretained time. A k value below 1 means the compound barely interacts with the column and elutes too quickly for reliable separation. Values between 2 and 10 generally give the best balance of good separation and reasonable run time. Values above 20 mean the compound is spending so long on the column that peaks become broad and flat.
Column Efficiency
Column efficiency is expressed as the number of theoretical plates, abbreviated N. A higher plate count means narrower, sharper peaks and better separation. You calculate N from a chromatogram by measuring a peak’s retention time and its width: N equals 16 times the retention time squared, divided by the peak width squared (measured at the base). Typical HPLC columns produce 10,000 to 20,000 theoretical plates, though this varies with column length and particle size.
The related concept of plate height (H) is simply the column length divided by N. A smaller plate height means more efficient separation per unit length of column. When comparing columns of different lengths, plate height is more useful than plate count because it normalizes for size.
Signal-to-Noise Ratio
At very low concentrations, peaks become hard to distinguish from baseline noise. The signal-to-noise ratio (S/N) compares the height of a peak to the height of the random fluctuations in the baseline. Two critical thresholds are widely used: a S/N of 3 is considered the limit of detection (LOD), meaning you can say a compound is present but can’t reliably measure how much. A S/N of 10 is the limit of quantification (LOQ), meaning the peak is large enough relative to noise that you can trust a concentration measurement.
If your peaks of interest are hovering near these thresholds, the chromatogram is telling you that you’re working at the edge of what the method can measure. You’d need to either concentrate the sample, inject more volume, or switch to a more sensitive detector.
Common Baseline Problems
A clean, flat baseline is the foundation of reliable chromatography. When the baseline misbehaves, it affects every measurement on the chromatogram.
Baseline drift, where the baseline slowly rises or falls over the course of the run, often comes from an aging detector lamp, temperature differences between the column and detector, or impurities that have accumulated on the column and are gradually leaching off. During gradient runs (where the mobile phase composition changes over time), some drift is normal because the changing solvent mixture absorbs differently at the detection wavelength.
Periodic baseline fluctuations, a rhythmic waviness, usually point to pump problems: leaking seals, worn pistons, check valve issues, or poorly degassed solvents with trapped air bubbles.
Ghost peaks are peaks that appear even when you inject a blank sample with no analyte. They show up because contaminants have built up on the column from previous injections and are washing off, or because bacterial or algal growth in the mobile phase reservoir is producing detectable byproducts. Running a blank injection before your samples helps you spot ghost peaks so you don’t mistake them for real compounds.
Reading a Chromatogram Step by Step
When you first look at a chromatogram, start with the baseline. Is it flat and stable, or noisy and drifting? A bad baseline undermines everything else. Next, count the peaks. Each well-resolved peak is a compound in your sample. Check whether peaks return to baseline between them; if they don’t, your quantification of those compounds will be less accurate.
For each peak, note the retention time and compare it to your standards to identify the compound. Check the peak shape for tailing or fronting, which could indicate system problems or overloading. Then look at the peak areas reported by your integrator. Compare them to your calibration curve to get concentrations. If you’re using an internal standard, calculate the area ratios before converting to concentration.
Finally, check whether the integration lines drawn by the software make sense. Automated integrators sometimes draw peak boundaries incorrectly, especially with overlapping peaks, shoulders, or noisy baselines. Manual adjustment of integration parameters is one of the most common tasks in chromatography, and getting it right is often the difference between accurate and misleading results.