🔬 Adsorption Rate Calculator

Determine Phage Adsorption Rate Constants from Free-Phage Decline Data

by Stephen T. Abedon Ph.D. (abedon.1@osu.edu)

phage.org | phage-therapy.org | biologyaspoetry.org | abedon.phage.org | google scholar

Jump to:   📊 Rate Constant Calculator  |  📂 Examples  |  🔄 Unit Converter & Visualizer  |  ⚗️ k and Killing  |  📖 Background & Methods  |  🧮 More Calculators

What is the adsorption rate constant (k)? The adsorption rate constant k describes the per-bacterium, per-unit-time rate at which free phages are removed from suspension by adsorption. Free phage titer declines exponentially at a rate determined by k and bacterial concentration, appearing as a straight line on a semi-log plot. Units are mL min⁻¹.

Enter time-series free-phage data to calculate k by linear regression. The Unit Converter tab interconverts common k unit formats.

To cite this tool: Abedon, S.T. (2026). Adsorption Rate Calculator. adsorption.phage.org. DOI: 10.5281/zenodo.21132369

adsorption.phage.org  ·  Abedon’s Books  ·  DOI: 10.5281/zenodo.21132369

How can I improve this page?  contact: adsorption@phage.org

Step 1 — Experimental Parameters

The t=0 titer is P₀ — a defined reference value, not an independent measurement. Including it forces the regression through the origin and biases k downward. Leave checked unless you have a specific reason to include it.
New here? See how the calculator works with real published data:

Step 2 — Enter Phage Titer Data ↓ Skip to data entry

What to enter: Free phage titers (PFU/mL) at successive time points from your adsorption assay — these should be measured from the supernatant after low-speed centrifugation (to pellet bacteria) or after chloroform treatment (to lyse phage-adsorbed bacteria). The time-zero value is P₀, the starting free-phage titer.

Three ways to enter data — use whichever is most convenient:
  1. Upload a spreadsheet — drag and drop or click the upload zone below. Your file should have at minimum two columns: one for time and one for free-phage titer. The first row may be a header (e.g., "Time (min)", "PFU/mL") or raw numbers — both work. Column order does not matter; you will select which column is which after upload. Accepted formats: .xlsx, .xls, .csv, .tsv, .txt.
  2. Paste from a spreadsheet or table — click Paste Data to open a text box and paste two columns (time and titer, one row per time point) copied directly from Excel, a text editor, or similar. Values may be tab-, comma-, or space-separated.
  3. Add rows manually — click + Add Row to type values one at a time.

After entering data, individual rows can be excluded from the regression using the checkboxes — useful for dropping late time points that deviate from linearity (e.g., due to bacterial growth or virion release). Excluded points still appear on the graph for reference. Rows can also be deleted individually using the ✕ button on each row.

Note on t=0: The "Exclude t=0 from regression" option in Step 1 is checked by default. See Step 1 for details.
Load Monophasic Example loads a simulated single-phase dataset (phage T4-like kinetics, k ≈ 2.5 × 10⁻⁹ mL min⁻¹, N = 2 × 10⁸ cells/mL) to demonstrate the calculator. Load Biphasic Example loads a two-phase dataset where k drops partway through the experiment. Both replace any data currently in the table. When 🎲 Randomise with noise is checked (the default), realistic Poisson-distributed scatter is applied to the titer values each time. For each time point the simulator independently draws a random expected plate count between 40 and 400 — the conventional countable range — and samples from a Poisson distribution with that count. This gives a coefficient of variation of ~5–16% per point, matching a well-run assay where the experimenter chooses dilutions to land within the countable window throughout. Each load gives a different random draw. Uncheck to see the exact theoretical values. Note: the 40–400 range is itself a simplification — the upper limit reflects plaque size and plating surface area rather than statistics, and the true range varies by phage and conditions (Abedon & Katsaounis, 2021, 10.1007/978-3-319-41986-2_17).
📂

Click to upload or drag & drop a spreadsheet here

Accepts .xlsx, .xls, .csv, .tsv, .txt — needs at minimum a time column and a free-phage titer column

# Time Free Phage Titer (PFU/mL) Exclude from fit? Ignore row? Delete row

Step 3 — Calculate

Published Adsorption Curve Examples

The datasets below are digitized or obtained from published adsorption experiments. Click Load into Calculator on any card to transfer the data directly to the Rate Constant Calculator tab, where it will be graphed and analysed automatically. No judgement is made here about data quality or curve shape — all interpretation is left to the calculator and to you.

Note on bacterial concentration (N): Some datasets do not specify N in the digitized figure. Where N is not known, the calculator will load the data and plot the curve, but you must enter N manually in Step 1 before calculating k. Where N is known it will be pre-filled automatically.
Loading examples…

Adsorption Rate Constant Unit Converter

Enter a k value and select its input unit, then select the output unit you want. All conversions are also shown as cards below. The standard unit in the phage literature is mL min⁻¹ (volume cleared per bacterium per minute). Note: 1 mL = 1 cm³, so mL and cm³ units are numerically identical.

Comparative Adsorption Visualizer

Visualize how different k values translate into free-phage decline over time. Click a preset button below or edit the k input fields that appear beneath it (up to 5 values, in mL min⁻¹). Set the bacterial concentration and duration, then toggle between semi-log (straight lines — the correct representation) and linear-linear (curves — how data appears when plotted incorrectly).
Graph scale:

k and Killing

What this tab does: Given an adsorption rate constant k, this tab shows how fast bacteria are expected to decline at various phage concentrations — helping you design a killing experiment and plan how to separate phages from bacteria before plating. It also lets you calculate k from bacterial survival data (the mirror image of the free-phage assay in the Rate Constant Calculator tab).

The math: When P₀ > B₀ and t is short, phage concentration stays somewhat constant and bacterial loss follows pseudo-first-order kinetics: B(t) = B₀ · ek·P₀·t. The slope of ln(B/B₀) vs t equals −k·P₀, so k = −slope / P₀. When B₀ is not negligible, phage are depleted as adsorption proceeds. The predicted killing curves in Section 1 account for this by numerically integrating the coupled equations dB/dt = −k·P·B and dP/dt = −k·P·B, where B is uninfected bacteria and P is free phage, so that both populations decline in tandem as adsorption proceeds. The regression calculator in Section 3 uses the pseudo-first-order approximation (constant P₀), which is accurate when P₀ ≫ B₀.

Key assumptions: Short assay (≤5 min) so bacterial replication, spontaneous death, and phage replication are negligible; single-hit kinetics; EOP near 1.

1. Predicted Killing Curves

Theory: A Rapid k Assessment

A rapid means of adsorption rate constant assessment can be achieved by measuring rates of bacterial killing at high phage concentration. The assay assumes an efficiency of plating (EOP) near 1 — that is, each phage-infected bacterium gives rise to a plaque, so CFU decline reflects adsorption events directly.

At P₀ = 5×10⁷ PFU/mL and B₀ = 10⁷ CFU/mL over 5 minutes:

Fast phage (k ≈ 10⁻⁸)
~92% killed
Clear, measurable decline
Moderate (k ≈ 10⁻⁹)
~22% killed
Detectable with good counts
Slow phage (k ≈ 10⁻¹⁰)
~2.4% killed
Within background noise — not detectable

A measurable decline in bacterial numbers at these conditions is indicative of an adsorption rate constant in the range of roughly 10⁻⁹ mL min⁻¹ or greater. No detectable decline suggests k ≪ 10⁻⁹ mL min⁻¹ (see EOP note below).

Plating note: Starting at B₀ = 10⁷ CFU/mL, a 10⁵-fold dilution gives ~100 CFU/plate. At P₀ = 5×10⁷ PFU/mL the same dilution leaves ~500 phages/plate, which should have no post-plating impact on colony counts.

EOP note: If EOP < 1, you cannot be sure of the killing phage titer and therefore cannot reliably assess adsorption rates from bacterial killing. Low EOP with high killing capacity will result in an overestimation of adsorption rates, since the perceived phage titer (based on plaque counts) is lower than the actual killing titer. At the extreme, a killing-positive but replication-negative phage has an EOP approaching zero but killing ability upon adsorption approaching 100%, resulting in a measured k that approaches infinity.

Enter k and initial bacterial concentration below to see predicted survival curves at various phage concentrations. Use these to choose P₀ for your experiment and to estimate how much dilution is needed before plating.

Default: 2.5×10⁻⁹ mL min⁻¹ (Stent, 1963). Automatically updated when k is calculated in the Rate Constant Calculator tab.

2. Plating Guidance

To ensure reasonable accuracy, diluting must take place immediately at specific time points, such as at 5 min. The zero-min time point can be taken from a phage-free culture, making sure to take into account any dilution that would result from subsequent phage addition. To minimize the impact of diluting errors, determine the zero-min time point using at least three dilution series. For other time points, perform an initial 100-fold dilution to stop adsorptions and then perform multiple subsequent parallel dilutions (many dilution series rather than many platings from a single dilution series).

Dilute into buffer so as to temporarily halt bacterial replication. Consider using median rather than mean determinations for titer estimations (Abedon and Katsaounis, 2021 — 10.1007/978-3-319-41986-2_17).

It is also possible to separate free phages from bacteria via centrifugation (filtration is not recommended since it is the bacteria that remain on the filter) or to preferentially inactivate free phages by various means. Note, though, that phage-infected bacteria will still release new phages on plates, though at ~100 bacteria per plate, that impact should be negligible.

3. Calculate k from Bacterial Survival Data

Once you have measured B(t) at several time points, enter them below to calculate k. Do not include t=0 — B₀ is the reference value entered above.

Must be ≫ B₀.
Time B(t) (CFU/mL) ln(B/B₀) k from point (mL min⁻¹)

Theory: The Adsorption Rate Constant

The rate of phage adsorption to bacteria is governed by mass-action kinetics: the instantaneous rate at which free phages are lost from suspension is proportional to the product of phage concentration (P), bacterial concentration (N), and the adsorption rate constant (k). This gives the differential equation dP/dt = −kNP. When N is held approximately constant — as in a well-designed short adsorption assay — this integrates to the exponential decay expression used throughout this calculator.

ln(P/P₀) = −k · N · t

Rearranged:   k = −ln(P/P₀) / (N · t)

From slope:    slope of ln(P) vs. t = −k · N   →   k = −slope / N

Log₁₀ correction: if slope is measured from a log₁₀ plot, multiply by ln(10) ≈ 2.303 before dividing by N

The units of k are mL min⁻¹ (equivalent to cm³ min⁻¹). This reflects a "clearance" perspective: k describes the volume effectively swept clear of free phages by a single bacterium per unit time. Multiplying by N gives the first-order rate constant for free-phage loss (units: min⁻¹), and the reciprocal 1/(kN) is the mean free time — the average time a phage spends searching before it adsorbs.

Note that the rate at which an individual phage finds bacteria is determined by k × N, while the rate at which an individual bacterium acquires phages is determined by k × P. These two perspectives on the same constant are relevant to different practical questions — the former to free-phage clearance in adsorption assays, the latter to phage therapy dosing.

What Determines k? The Collision Kernel

From the physics of diffusion-driven particle collisions, k can be decomposed as:

k = S · C · f

where S is a measure of bacterial target size (proportional to cell radius R, such that S = 4πR), C is the virion diffusion constant (larger virions diffuse more slowly; higher medium viscosity reduces C), and f is the efficiency of adsorption given collision — the probability that a phagebacterium encounter actually results in irreversible attachment. The value of f reflects the density and affinity of phage receptor molecules on the bacterial surface.

In practice, k therefore tends to be larger for phages infecting bigger bacteria, for smaller (faster-diffusing) virions, and for phages with high receptor affinity. Measured values span roughly 10⁻⁷ to 10⁻¹¹ mL min⁻¹ across different phage–host pairs.

Why Semi-Log Graphing Is Essential

Because phage loss is exponential, plotting phage titers against time on a linear y-axis produces a sharply falling curve that quickly flattens near zero. On such a linear-linear plot it is nearly impossible to assess whether the decline is truly exponential, to determine the slope accurately, or to detect a change in adsorption rate. Plotting the same data with a logarithmic y-axis (semi-log or log-linear plot) converts the exponential decay into a straight line. The slope of that line is −kN, from which k follows directly after dividing by N. Non-linearities — whether from bacterial growth, virion release, phage aggregation, or a biphasic adsorption process — are far more visible on the semi-log scale. Despite this, linear-linear graphing remains common in the literature and is one of the most frequently cited methodological errors in adsorption studies.

Key insight: A straight line on a semi-log plot is the primary diagnostic that adsorption is following simple mass-action kinetics with constant k and N throughout the assay. Deviations from linearity should prompt investigation of experimental conditions and, where appropriate, restriction of the regression to the initial linear region.

Biphasic Adsorption

Not all phage populations adsorb at a single constant rate. A biphasic adsorption curve arises when a fraction of phages adsorbs rapidly while the remainder adsorbs more slowly — or not at all. On a semi-log plot this appears as an initial steep linear decline followed by a shallower (or flat) second phase. On a linear-linear plot the two phases may be nearly invisible, making semi-log presentation critical for detecting this phenomenon.

Possible causes include phage population heterogeneity (e.g., a fraction that has lost tail fibers), a subpopulation of resistant or non-susceptible bacteria, reversible phage aggregation, or saturation of bacterial receptor sites at high multiplicities. The Load Biphasic Example button in Step 2 loads a simulated dataset illustrating this pattern, based on the example values used by Abedon (2023) (k dropping from 2.5 × 10⁻⁹ to 2.5 × 10⁻¹⁰ mL min⁻¹ at a breakpoint). When analyzing biphasic data, restrict your regression to the initial linear phase and exclude later points manually using the checkboxes.

R² and the Correlation Coefficient

R² (the coefficient of determination) equals the square of the Pearson correlation coefficient r: R² = r². The Pearson r ranges from −1 to +1 and measures the strength and direction of the linear relationship between ln(P) and time t; R² then measures the proportion of variance in ln(P) explained by that linear relationship, ranging from 0 to 1. For a declining adsorption curve, r will be negative, so it is conventional to report R² rather than r. An R² of 0.98 corresponds to r = −0.990; an R² of 0.99 corresponds to r = −0.995. Values below about 0.98 suggest the data depart meaningfully from a straight line on the semi-log plot.

Methods: Separating Free Phages from Adsorbed Phages

The central experimental requirement for an adsorption assay is the ability to measure free-phage titers independently of phages that have adsorbed to bacteria. Three approaches are widely used, each with specific limitations:

Regardless of method, assay duration should generally not exceed 10 minutes. Longer assays allow bacterial growth (which increases N and accelerates adsorption over time, causing downward curvature on the semi-log plot) and risk virion release from lysing cells (which artificially inflates free-phage counts, causing upward curvature).

Practical Notes

  • ⚠ Do not include t=0 as a data point. The t=0 titer is P₀ — a defined reference value, not an independent measurement. Including it in the regression forces the fit through the origin and biases k downward. This is an extremely common error. Enter only the time points at which free-phage titer was actually assayed.
  • Use natural log (ln) in calculations; use log₁₀ when displaying curves, but apply the 2.303 correction factor to recover k.
  • Use multi-point regression rather than a two-point endpoint calculation; this reveals non-linearities that a single endpoint cannot detect.
  • Aim for ≥4 time points spanning roughly one order of magnitude in titer decline.
  • Determine N independently by plate count or calibrated OD immediately before the assay.
  • Keep the phage-to-bacterium ratio (MOI) low to minimize multiple adsorptions per cell.

How to Cite This Tool

Abedon, S.T. (2026). Adsorption Rate Calculator. adsorption.phage.org. DOI: 10.5281/zenodo.21132369

References

Much of the information in this calculator can be found in the following references. Please cite this tool as: Abedon, S.T. (2026). Adsorption Rate Calculator. adsorption.phage.org. DOI: 10.5281/zenodo.21132369.

  • Hyman, P. and Abedon, S.T. (2009). Practical methods for determining phage growth parameters. Methods in Molecular Biology 501:175–202. 10.1007/978-1-60327-164-6_18
  • Dennehy, J.J. and Abedon, S.T. (2021). Adsorption: phage acquisition of bacteria. In: Bacteriophages: Biology, Technology, Therapy. Springer Nature Switzerland AG. pp. 93–117. 10.1007/978-3-319-40598-8_2-1
  • Abedon, S.T. (2023). Bacteriophage adsorption: likelihood of virion encounter with bacteria and other factors affecting rates. Antibiotics 12:723. 10.3390/antibiotics12040723
  • Abedon, S.T. (2023). Schlesinger nailed it! Assessing a key primary pharmacodynamic property of phages for phage therapy: virion encounter rates with motionless bacterial targets. Drugs and Drug Candidates 2:673–688. 10.3390/ddc2030034
  • Stent, G.S. (1963). Molecular Biology of Bacterial Viruses. WH Freeman, San Francisco.
  • Abedon, S.T. and Katsaounis, T.I. (2021). Basic phage mathematics. In: Bacteriophages: Biology, Technology, Therapy. Springer Nature Switzerland AG. 10.1007/978-3-319-41986-2_17

🧮 Phage Biology, Phage Therapy, and Teaching Calculators

Loading calculators…