Pseudo-First-Order Approximation

Why an Excess Reactant Can Simplify a Rate Law

For an elementary bimolecular reaction, the mass-action rate is proportional to the product of the two reactant concentrations. When one reactant is present in much greater quantity than the other, its concentration may change little over the time interval of interest. Treating that concentration as constant converts the bimolecular rate law into a first-order law in the limiting reactant. This is the pseudo-first-order approximation: the reaction remains bimolecular in its underlying mechanism, but its observed rate has first-order form under a specified condition.

The approximation is useful for simplifying calculations and interpreting time courses, but “in excess” is not by itself a guarantee of accuracy. The key question is how much the excess reactant is depleted over the interval being modeled. The comparison below uses the exact solution of the same irreversible reaction to make that question quantitative.

Model and Assumptions

Consider the elementary reaction $A+B\rightarrow C$, with one molecule of $A$ reacting with one molecule of $B$. Let $A(t)$, $B(t)$, and $C(t)$ denote their concentrations, and let $A_0=A(0)$, $B_0=B(0)$. Under mass-action kinetics, the reaction rate is $kA(t)B(t)$. Assume a well-mixed, constant-volume system, constant temperature, a time-independent rate constant $k$, and no inflow, outflow, or competing reaction. The stoichiometry then gives the ODEs for $A$ and $B$.

\frac{dA}{dt}=-kAB,\qquad \frac{dB}{dt}=-kAB

The rate constant $k$ has units of concentration$^{-1}$ time$^{-1}$, so $kAB$ has units of concentration per time. Subtracting the two ODEs shows that $B(t)-A(t)=B_0-A_0$ remains constant. Write $D=B_0-A_0$. For the usual excess-reactant case, $D>0$: $B$ starts above $A$ and remains above it as both are consumed.

Definition: If $B$ is in sufficient excess that its concentration can be treated as approximately constant over the interval of interest, replace $B(t)$ by $B_0$ in the rate law for $A$. The resulting pseudo-first-order model is $$ \frac{dA}{dt}\approx-kB_0A=-k_{\mathrm{obs}}A,\qquad k_{\mathrm{obs}}=kB_0 $$ Here $k_{\mathrm{obs}}$ is the apparent first-order rate constant, with units of inverse time. The approximation predicts $A(t)\approx A_0e^{-kB_0t}$.

Deriving the Exact Comparison

To assess the approximation, retain the depletion of $B$. Since $B=A+D$, the exact ODE for $A$ is $dA/dt=-kA(A+D)$. When $D>0$, separation of variables gives an exact solution.

\frac{1}{A(A+D)}=\frac{1}{D}\left(\frac{1}{A}-\frac{1}{A+D}\right)

Integrating from $A_0$ at time zero to $A(t)$ at time $t$, and using $A_0+D=B_0$, gives

\frac{1}{D}\ln\left(\frac{A(t)}{A(t)+D}\right) -\frac{1}{D}\ln\left(\frac{A_0}{B_0}\right)=-kt

Exponentiating and solving for $A(t)$ yields the exact concentration trajectory. This expression lets us check whether holding $B$ at $B_0$ is adequate for a particular initial condition and time window.

A_{\mathrm{exact}}(t)= \frac{A_0D e^{-kDt}}{B_0-A_0e^{-kDt}}, \qquad D=B_0-A_0>0

The pseudo-first-order approximation uses $B_0$ as its effective partner concentration throughout, whereas the exact trajectory uses the decreasing $B(t)$. Initially, both models have the same rate of loss of $A$, $-kA_0B_0$. Later, if $B$ has fallen appreciably, the exact reaction slows relative to the approximation. The derivation also shows why the two trajectories need not remain close indefinitely, even when $A_0/B_0$ is initially small.

Worked Example: Small Depletion With a Large Initial Excess

Suppose the hypothetical reaction has $A_0=0.200\ \mathrm{mM}$, $B_0=10.0\ \mathrm{mM}$, and $k=0.0400\ \mathrm{mM}^{-1}\mathrm{min}^{-1}$. Compare the two predictions at $t=2.00\ \mathrm{min}$. The assumptions are those of the stated model: a single irreversible 1:1 step, constant volume and temperature, and no other sources or sinks.

First, $D=B_0-A_0=9.80\ \mathrm{mM}$, so the exact solution has exponent $-kDt=-(0.0400)(9.80)(2.00)=-0.784$. Substituting into the exact formula gives

A_{\mathrm{exact}}(2.00) =\frac{(0.200)(9.80)e^{-0.784}}{10.0-(0.200)e^{-0.784}} \approx 0.09031\ \mathrm{mM}

The pseudo-first-order prediction uses $k_{\mathrm{obs}}=kB_0=(0.0400)(10.0)=0.400\ \mathrm{min}^{-1}$. Therefore,

A_{\mathrm{approx}}(2.00) =(0.200)e^{-(0.400)(2.00)} \approx 0.08987\ \mathrm{mM}

The approximation is lower by about $0.09031-0.08987=0.0004478\ \mathrm{mM}$, approximately $0.496\%$ of the exact concentration. This small discrepancy is consistent with the small initial ratio $A_0/B_0=0.0200$. In this model, the amount of $B$ consumed by any time cannot exceed $A_0$, so its fractional depletion is at most $A_0/B_0=2.00\%$.

Depletion Gives a Direct Accuracy Check

The stoichiometry supplies a simple screening calculation before solving either ODE. At most $A_0$ of $B$ can be consumed, because each reaction event consumes one $A$ and one $B$. Thus the fractional decrease in $B$, relative to its initial concentration, satisfies

\frac{B_0-B(t)}{B_0}=\frac{A_0-A(t)}{B_0} \leq \frac{A_0}{B_0}

If $A_0/B_0$ is small, then $B(t)$ stays close to $B_0$ in absolute fractional terms, which supports the pseudo-first-order assumption. This bound is a useful diagnostic, not a universal guarantee of a specified error in every predicted quantity. In particular, the approximate and exact decay rates differ: the exact rate coefficient for $A$ is $kB(t)$, which declines from $kB_0$. Small differences in rate can accumulate over time, and relative differences between two very small late-time concentrations can be large.

A more explicit time-course diagnostic is the ratio $A_{\mathrm{exact}}(t)/A_{\mathrm{approx}}(t)$. A ratio close to one means the two concentration predictions are close in relative terms at that time. It should be evaluated over the time interval relevant to the model’s use, rather than only at the initial instant. If the approximation informs a threshold crossing or a half-time, compare that quantity directly as well.

Worked Example: Substantial Depletion and a Half-Time Error

Consider a second hypothetical reaction with $A_0=2.00\ \mathrm{mM}$, $B_0=2.50\ \mathrm{mM}$, and $k=0.200\ \mathrm{mM}^{-1}\mathrm{min}^{-1}$. The excess is modest: $A_0/B_0=0.800$, so $B$ could lose as much as $80.0\%$ of its starting concentration. The model assumptions are unchanged, but the constancy assumption is now suspect.

For the pseudo-first-order model, the half-time is obtained by setting $A(t)=A_0/2$ in $A_0e^{-kB_0t}$. Thus

t_{1/2,\mathrm{approx}}=\frac{\ln 2}{kB_0} =\frac{0.693147}{(0.200)(2.50)} \approx 1.386\ \mathrm{min}

For the exact model, set $A_{\mathrm{exact}}(t)=A_0/2$, with $D=B_0-A_0$. Rearranging the exact solution gives $e^{-kDt}=B_0/(2B_0-A_0)$, and hence

t_{1/2,\mathrm{exact}} =\frac{\ln((2B_0-A_0)/B_0)}{k(B_0-A_0)} =\frac{\ln(3.00/2.50)}{(0.200)(0.500)} \approx 1.823\ \mathrm{min}

Using the unrounded values, the approximate-minus-exact difference is about $-0.436921\ \mathrm{min}$ to six decimal places, or about $-0.437\ \mathrm{min}$ to three decimal places. The pseudo-first-order half-time is therefore about $24.0\%$ shorter than the exact half-time. In this case, treating $B$ as constant materially overstates how quickly $A$ is removed.

Worked Example: A Small Ratio and a Half-Time Check

For a third hypothetical case, let $A_0=0.500\ \mathrm{mM}$, $B_0=10.0\ \mathrm{mM}$, and $k=0.0600\ \mathrm{mM}^{-1}\mathrm{min}^{-1}$. Here $A_0/B_0=0.0500$, so no more than $5.00\%$ of $B_0$ can be consumed. Calculate both half-times to see how this small depletion bound translates for this endpoint.

The pseudo-first-order half-time is

t_{1/2,\mathrm{approx}} =\frac{\ln 2}{(0.0600)(10.0)} \approx 1.155245\ \mathrm{min}

For the exact half-time, $D=10.0-0.500=9.50\ \mathrm{mM}$, so

t_{1/2,\mathrm{exact}} =\frac{\ln((20.0-0.500)/10.0)}{(0.0600)(9.50)} =\frac{\ln(1.95)}{0.570} \approx 1.171630\ \mathrm{min}

The approximate-minus-exact difference, computed from the unrounded values, is about $-0.016385\ \mathrm{min}$, or $-0.0164\ \mathrm{min}$ to three significant figures. Relative to the exact half-time, the approximation is shorter by approximately $(0.016385/1.171630)\times100\%=1.398\%$. This endpoint is predicted closely here, but the result is specific to the chosen concentrations, rate constant, and quantity being compared.

Choosing and Checking the Approximation

There are three practical approaches for this reaction. The pseudo-first-order model is simplest and is appropriate when the excess-reactant concentration changes negligibly over the interval of interest. The exact mass-action solution is preferable when the model assumptions hold and depletion matters; it requires no numerical approximation for this two-reactant system. A numerical solution of the original ODEs is useful when additional states, inputs, or reaction steps make an exact solution unavailable, but numerical integration does not remove the need to choose a valid model.

A useful workflow is to calculate the depletion bound $A_0/B_0$, then compare the approximate prediction with an exact or numerically integrated mass-action prediction at times relevant to the application. Check the concentration trajectory and any derived quantities of interest. If predictions differ beyond the acceptable tolerance, retain $B$ as a dynamic state rather than fixing it at $B_0$.

The exact expression above assumes $B_0>A_0$. If $B_0=A_0$, then both concentrations remain equal and satisfy $dA/dt=-kA^2$, giving $A(t)=A_0/(1+kA_0t)$, not an exponential. If $B_0<A_0$, then $B$, rather than $A$, is the limiting reactant; holding $B$ constant is not justified by the usual excess-reactant argument. These cases are reminders that the approximation depends on the stoichiometry and initial conditions, not merely on the appearance of a first-order formula.

Common Mistakes and Limitations

Calling the underlying reaction first-order. The elementary step remains bimolecular. It is the reduced rate law for $A$, under a nearly constant $B$, that has first-order form.
Using $k$ as the apparent first-order constant. The apparent constant is $k_{\mathrm{obs}}=kB_0$, with units of inverse time; $k$ retains units of concentration$^{-1}$ time$^{-1}$.
Assuming “large excess” means exact constancy. The assumption is an approximation. Check the possible depletion $A_0/B_0$ and compare predictions over the interval that matters.
Checking only the initial rate. Both models have the same initial rate, but the exact rate coefficient $kB(t)$ decreases as $B$ is consumed. Compare trajectories or the final quantity of interest.
Using this reduction when there are other processes changing $B$. Inflow, outflow, or additional reactions can change $B$ independently of $A$. The simple depletion bound and exact solution then do not apply without modifying the model.

The pseudo-first-order approximation is therefore a controlled simplification, not a change in reaction mechanism. Its central justification is that the excess reactant remains nearly constant, and its central diagnostic is whether that condition holds over the time and concentration range used in the analysis.

Check Your Understanding

Answer each question, then compare with the brief solutions below.

For $A+B\rightarrow C$, with $k=0.0300\ \mathrm{mM}^{-1}\mathrm{min}^{-1}$ and $B_0=8.00\ \mathrm{mM}$, what is $k_{\mathrm{obs}}$ under the pseudo-first-order approximation?
If $A_0=0.300\ \mathrm{mM}$ and $B_0=6.00\ \mathrm{mM}$, what is the maximum fractional depletion of $B$ in the stated 1:1 reaction?
Why can the exact and pseudo-first-order models have the same initial rate but different later-time predictions?
For $A_0=B_0$, explain why the ordinary excess-reactant argument does not support a pseudo-first-order model for $A$.

Solutions. (1) $k_{\mathrm{obs}}=kB_0=(0.0300)(8.00)=0.240\ \mathrm{min}^{-1}$. (2) The bound is $A_0/B_0=0.300/6.00=0.0500$, or $5.00\%$. (3) Initially $B(0)=B_0$, so both use the same rate $-kA_0B_0$; later, the exact model has a reduced $B(t)$ and hence a lower rate coefficient $kB(t)$. (4) Neither reactant is initially in excess, and their concentrations fall together; in fact $A=B$ and $dA/dt=-kA^2$, which is not a first-order exponential law.