Exponentials and logarithms (AS Unit 1: Pure Mathematics A) — WJEC A-Level Mathematics
Exponentials and logarithms (AS Unit 1: Pure Mathematics A) explained
Exponentials and logarithms (AS Unit 1: Pure Mathematics A) covers 13 specification points, including Know and use the function aˣ and its graph,… and Know and use the function eˣ and its graph..
Read the full explanation
Open a point below for its explanation and revision guidance.
Know and use the function aˣ and its graph, where a is positive.
The exponential function f(x) = aˣ with base a > 0 is defined for all real x, with range f(x) > 0. When a > 1 the function shows exponential growth: it is strictly increasing, passes through the y-intercept (0, 1) and the point (1, a). As x → −∞, aˣ approaches zero from above, so the line y = 0 is a horizontal asymptote; as x → +∞, aˣ increases without bound. When 0 < a < 1 the function shows exponential decay: strictly decreasing, passing through (0, 1), with horizontal asymptote y = 0 as x → +∞. The graph never touches or crosses the x-axis. In applications, aˣ models unconstrained multiplication such as doubling populations (2ˣ). Transformations behave as expected: y = aˣ + k shifts the horizontal asymptote to y = k, and y = a^(x − h) shifts the graph horizontally.
Know and use the function eˣ and its graph.
The natural exponential function f(x) = eˣ uses Euler's constant e ≈ 2.71828 as its base. It is unique in calculus because its rate of change is identically equal to its function value: d/dx(eˣ) = eˣ, meaning the gradient of the tangent at any point (x, y) equals y itself. In particular, the gradient at the y-intercept (0, 1) is exactly 1. The domain is all real numbers x ∈ ℝ, and the range is strictly positive reals, y > 0. The graph passes through (0, 1) and (1, e). As x → −∞, eˣ → 0, so the line y = 0 is a horizontal asymptote. As x → +∞, eˣ grows faster than any polynomial function. Its inverse function is the natural logarithm y = ln x, whose graph is the reflection of y = eˣ in the line of symmetry y = x.
Know that the gradient of eᵏˣ is equal to keᵏˣ and hence understand why the exponential model is suitable in many applications.
Applying the chain rule to the natural exponential function yields the fundamental derivative d/dx(e^(kx)) = k e^(kx). This mathematical property states that the instantaneous rate of change of y = A e^(kx) is directly proportional to the current value of y, satisfying the first-order differential equation dy/dx = ky. When k > 0, the rate of change is positive and increases in proportion to y, accurately modelling unrestricted biological population growth, viral spread, or compound interest. When k < 0, the rate of change is negative and slows down as y approaches zero, accurately modelling radioactive decay, capacitor discharge, or drug elimination from the bloodstream. This intrinsic proportionality makes exponential models natural descriptors of real-world dynamics.
Know and use the definition of logₐ x as the inverse of aˣ , where a is positive and x ≥ 0.
A logarithm is the inverse of an exponential function. By definition, y = logₐ x is equivalent to aʸ = x, where the base satisfies a > 0 and a ≠ 1, and the argument satisfies x > 0. The logarithm logₐ x answers: to what power must base a be raised to give x? From this inverse relationship, logₐ(aˣ) = x for all real x, and a^(logₐ x) = x for x > 0. Special values: logₐ 1 = 0 since a⁰ = 1, and logₐ a = 1 since a¹ = a. The graph of y = logₐ x has domain x > 0, range y ∈ ℝ, x-intercept (1, 0), and vertical asymptote x = 0 approached as x → 0⁺. Base 10 is written lg x or log x; base e is ln x. Note: the specification statement writes x ≥ 0, but logₐ 0 is undefined, so the true domain is x > 0.
Know and use the function lnx and its graph.
The natural logarithm function y = ln x is defined as the logarithm to base e, so y = ln x ⇔ eʸ = x. Its domain is restricted to strictly positive real numbers, {x ∈ ℝ : x > 0}, and its range is all real numbers ℝ. The graph passes through the x-intercept at (1, 0) and through the point (e, 1). The y-axis (x = 0) is a vertical asymptote, with ln x → −∞ as x approaches 0 from the right (x → 0⁺). The function is continuous and strictly increasing, but its rate of growth slows as x increases, having derivative d/dx(ln x) = 1/x for x > 0. It is concave downwards everywhere (d²y/dx² = −1/x² < 0). Transformations such as y = ln(x − a) shift the vertical asymptote to x = a, restricting the domain to x > a.
Know and use lnx as the inverse function of eˣ.
The functions f(x) = eˣ and g(x) = ln x are mutual inverse functions. This means that composing them in either order returns the input variable: ln(eˣ) = x for all real x ∈ ℝ, and e^(ln x) = x for all x > 0. Geometrically, the graph of y = ln x is the reflection of the graph of y = eˣ in the diagonal line of symmetry y = x. The domain of eˣ (ℝ) becomes the range of ln x, and the range of eˣ (y > 0) becomes the domain of ln x (x > 0). This inverse relationship is the primary algebraic tool for unlocking variables trapped in exponents or logarithms: solving an exponential equation e^(2x) = 5 requires taking natural logarithms, 2x = ln 5; conversely, solving ln(3x + 1) = 4 requires taking exponentials, 3x + 1 = e⁴.
Understand and use the laws of logarithms.
Logarithm laws derive directly from the laws of indices. For any valid base a > 0 (a ≠ 1) and positive arguments x, y > 0: the addition law states logₐ x + logₐ y = logₐ (xy), corresponding to aᵐ aⁿ = a^(m+n); the subtraction law states logₐ x − logₐ y = logₐ (x/y), corresponding to aᵐ / aⁿ = a^(m−n); and the power law states logₐ (xᵏ) = k logₐ x, corresponding to (aᵐ)ᵏ = a^(mk). Special identities include logₐ 1 = 0, logₐ a = 1, and the change of base formula logₐ x = (log_b x) / (log_b a). These laws allow complex multi-term logarithmic expressions to be condensed into a single logarithm to solve equations, or expanded into linear combinations of simpler logarithms for differentiation and parameter estimation.
logₐ x + logₐ y = logₐ (xy)
The product law of logarithms asserts that the sum of two logarithms sharing the same base equals the logarithm of their product: logₐ x + logₐ y = logₐ (xy), for positive arguments x > 0 and y > 0. Proof: let m = logₐ x and n = logₐ y, which means x = aᵐ and y = aⁿ. Multiplying these gives xy = aᵐ aⁿ = a^(m+n). Converting back to logarithmic form yields logₐ (xy) = m + n = logₐ x + logₐ y. In algebraic problem solving, this law combines split logarithmic terms into a single compact argument, enabling direct exponentiation. For instance, solving log₂ x + log₂ (x − 2) = 3 combines the LHS to log₂ (x(x − 2)) = 3, giving the quadratic x(x − 2) = 2³ = 8, from which x = 4 is retained and x = −2 rejected as an invalid argument.
logₐ x − logₐ y = logₐ (x/y) k logₐ x = logₐ (xᵏ) (including, for example k = –1, k = -½)
The quotient and power laws govern subtraction and scaling of logarithms for positive arguments x > 0, y > 0. The quotient law states logₐ x − logₐ y = logₐ (x/y), derived from index subtraction aᵐ / aⁿ = a^(m−n). The power law states k logₐ x = logₐ (xᵏ) for any rational k. When k = −1, −logₐ x = logₐ (x⁻¹) = logₐ (1/x), expressing the logarithm of a reciprocal as the negative logarithm. When k = 1/2, (1/2) logₐ x = logₐ (x^(1/2)) = logₐ (√x); when k = −1/2, (−1/2) logₐ x = logₐ (1/√x). In solving logarithmic equations, coefficients must be moved into the argument as powers before the addition or subtraction laws can be applied; for example, 2 log₃ x − log₃ 4 = log₃ (x²) − log₃ 4 = log₃ (x²/4).
Solve equations in the form aˣ = b.
Exponential equations of the form aˣ = b with a > 0 and b > 0 are solved by taking logarithms of both sides. Applying natural logarithms gives ln(aˣ) = ln b. Using the power law k ln x brings the exponent down: x ln a = ln b, yielding the exact solution x = (ln b) / (ln a). Equivalently, by definition of logarithms, x = logₐ b, which evaluates identically via the change of base formula. For equations with linear expressions in the exponent, such as a^(px + q) = b, taking logarithms produces (px + q) ln a = ln b, which is a linear equation in x. For equations where both sides feature exponential expressions with different bases, such as 2^(x+1) = 5^(2x−1), taking natural logs yields (x + 1) ln 2 = (2x − 1) ln 5, which rearranges to isolate x.
Use logarithmic graphs to estimate parameters in relationships of the form y = axⁿ and y = kbˣ , given data for x and y.
Logarithmic transformations linearise non-linear experimental data into the straight-line form Y = mX + c, enabling unknown parameters to be estimated from linear regression or line-of-best-fit gradients and intercepts. For a power law y = axⁿ, taking natural logarithms yields ln y = ln(axⁿ) = ln a + n ln x, or ln y = n(ln x) + ln a. Plotting ln y against ln x produces a straight line where gradient m = n and vertical intercept c = ln a, so a = eᶜ. For an exponential relationship y = kbˣ, taking natural logarithms yields ln y = ln(kbˣ) = ln k + x ln b, or ln y = (ln b)x + ln k. Plotting ln y against x produces a straight line where gradient m = ln b (so b = eᵐ) and vertical intercept c = ln k (so k = eᶜ). Base-10 logarithms lg can also be used.
Understand and use exponential growth and decay; use in modelling (examples may include the use of e in continuous compound interest, radioactive decay, drug concentration decay, exponential growth as model for population growth.)
Exponential models describe physical processes where the rate of change is proportional to the current amount, expressed as y(t) = y₀ e^(kt). When k > 0, the model represents exponential growth: initial value y₀ at t = 0 increases without bound, used in continuous compound interest A(t) = P e^(rt) and unchecked population growth P(t) = P₀ e^(kt). The doubling time T is found by setting y = 2y₀: 2y₀ = y₀ e^(kT) ⇒ T = (ln 2)/k. When k < 0 (often written y(t) = y₀ e^(−λt) with λ > 0), the model represents exponential decay: initial value y₀ decreases towards zero as t → ∞, used in radioactive decay and drug concentration clearance. The half-life t_(1/2) is found by setting y = 0.5y₀: t_(1/2) = (ln 2)/λ.
Consideration of limitations and refinements of exponential models.
Exponential models are assessed by comparing their predictions with physical reality over short, medium and long time horizons. In growth P(t) = P₀e^(kt), the model assumes unlimited resources, no competition and a constant growth rate. As t → ∞, P(t) → ∞, which is physically impossible; real populations face food scarcity and disease. A refinement suitable for AS is a bounded model such as P(t) = K/(1 + Ae^(−kt)), where P approaches a carrying capacity K asymptotically. In decay such as cooling θ(t) = θ₀e^(−kt), the prediction approaches zero, ignoring ambient room temperature θ_room; a baseline refinement gives θ(t) = θ_room + (θ₀ − θ_room)e^(−kt).
Your focus
- Sketch exponential functions y = aˣ for a > 1 and 0 < a < 1 with key features labelled.
- Identify the domain (ℝ), range (y > 0), and horizontal asymptote (y = 0) of exponential curves.
- Apply vertical and horizontal transformations to exponential functions.
Show all 39 objectives
- Define and sketch the natural exponential function y = eˣ and identify its key properties.
- Explain the unique calculus property that the derivative of eˣ is equal to itself.
- Relate y = eˣ to its inverse function y = ln x geometrically and algebraically.
- Differentiate exponential functions of the form y = e^(kx) to obtain dy/dx = k e^(kx).
- Explain why exponential functions are natural models for systems where rate of change is proportional to size.
- Apply exponential growth and decay models to real-world contexts including populations and cooling.
- Define logₐ x as the inverse of the exponential function aˣ.
- Convert equations fluently between logarithmic and exponential forms.
- Sketch logarithmic curves and identify their domain, range, and asymptotes.
- Sketch the graph of y = ln x, labelling its x-intercept and vertical asymptote.
- State the domain and range of the natural logarithm function and its transformations.
- Differentiate and manipulate expressions containing the natural logarithm.
- Demonstrate that y = eˣ and y = ln x are inverse functions algebraically and graphically.
- Apply natural logarithms to isolate variables in exponential equations.
- Apply exponential operations to solve logarithmic equations.
- Apply the product, quotient, and power laws to simplify logarithmic expressions.
- Combine multi-term logarithmic expressions into a single compact logarithmic term.
- Solve logarithmic equations by applying the laws of logarithms systematically.
- Prove the product law of logarithms from the laws of indices.
- Apply the product law to combine logarithmic terms into a single logarithm.
- Solve equations using the product law and verify roots against domain restrictions.
- Apply the quotient law to express differences of logarithms as a single logarithm of a quotient.
- Apply the power law to move coefficients inside arguments as powers, including fractional and negative exponents.
- Synthesise logarithm laws to condense multi-term expressions into a single solvable term.
- Solve simple exponential equations of the form aˣ = b using logarithms.
- Solve equations with differing bases by taking logarithms and rearranging linear terms.
- Express solutions to exponential equations in exact logarithmic form.
- Transform power laws y = axⁿ into linear relationships ln y = n ln x + ln a.
- Transform exponential relationships y = kbˣ into linear relationships ln y = x ln b + ln k.
- Estimate constants and powers from the gradient and vertical intercept of logarithmic graphs.
- Formulate exponential growth and decay functions of the form y = y₀ e^(kt).
- Calculate unknown parameters and rates from given contextual data points.
- Determine half-life and doubling time using natural logarithms.
- Identify the limitations of unconstrained exponential growth and decay models.
- Explain the physical factors that cause real-world processes to diverge from pure exponential curves.
- Propose mathematical refinements such as carrying capacities and baseline offsets.
Exponentials and logarithms (AS Unit 1: Pure Mathematics A) exam tips
Marking Points
- Sketching an exponential curve showing the correct growth or decay orientation according to whether a > 1 or 0 < a < 1.
- Marking the y-intercept at (0, 1) and identifying the point (1, a) on the curve.
- Drawing and labelling the horizontal asymptote along y = 0, showing the curve approaching but never touching or crossing it.
- Stating the domain (x ∈ ℝ) and range (f(x) > 0) of y = aˣ.
- Applying vertical and horizontal transformations, e.g. y = 2ˣ + 3 has horizontal asymptote y = 3.
- sketching y = eˣ showing the y-intercept at (0, 1) and horizontal asymptote y = 0
- stating that the derivative of eˣ is eˣ and explaining its geometric significance
- identifying the inverse relationship between y = eˣ and y = ln x reflected across y = x
- applying the chain rule to differentiate e^(kx), obtaining k e^(kx)
- demonstrating that dy/dx = ky confirms rate of change is directly proportional to y
- explaining why exponential growth (k > 0) or decay (k < 0) suits contextual applications
- Converting between logarithmic form y = logₐ x and exponential form aʸ = x, for example rewriting log₂ 8 = 3 as 2³ = 8.
- Stating the domain restriction x > 0 and base conditions a > 0, a ≠ 1, explaining that logₐ 0 and logₐ(negative) are undefined in the reals.
- Sketching y = logₐ x with x-intercept (1, 0), vertical asymptote x = 0, and correct behaviour for a > 1 versus 0 < a < 1.
- Using inverse properties logₐ(aˣ) = x and a^(logₐ x) = x, and special values logₐ 1 = 0, logₐ a = 1, to simplify expressions.
- Solving a simple logarithmic equation such as log₂(x − 1) = 3 by converting to 2³ = x − 1, giving x = 9, then checking x > 0.
- sketching y = ln x with x-intercept (1, 0) and vertical asymptote x = 0
- stating the domain x > 0 and range y ∈ ℝ of the natural logarithm function
- applying horizontal or vertical translations to y = ln x and identifying new asymptotes
- applying the inverse identities ln(eˣ) = x and e^(ln x) = x to solve equations
- sketching y = eˣ and y = ln x on the same axes showing reflection in the line y = x
- deducing domain and range correspondences between the inverse function pair
- combining logarithmic sums and differences into a single logarithm
- applying the power law to bring coefficients into exponents or vice versa
- applying the change of base formula or base identity logₐ a = 1
- applying logₐ x + logₐ y = logₐ (xy) to combine terms into a single logarithm
- removing logarithms by converting the single logarithmic equation into index form
- solving the resulting quadratic or polynomial equation
- checking candidate solutions and rejecting values that make arguments non-positive
- applying the power law to move coefficients into arguments as exponents
- applying the quotient law logₐ x − logₐ y = logₐ (x/y) to combine terms into a single fraction
- evaluating cases with negative or fractional coefficients like k = −1 or k = −1/2
- converting the resulting single logarithm to index form and solving for x
- taking logarithms (ln or log) of both sides of the exponential equation
- applying the power law to bring exponents down as linear multipliers
- expanding brackets and grouping terms in x on one side of the equation
- factorising out x and stating the exact solution in terms of logarithms
- applying logarithms to both sides and linearising into the form Y = mX + c
- calculating the gradient m of the straight-line graph from given coordinates
- determining the vertical intercept c from the graph or through substitution
- reversing logarithmic transformations to calculate the original parameters (e.g. a = eᶜ)
- substituting t = 0 to identify the initial value y₀ in the exponential model
- substituting given boundary coordinates to find the growth or decay constant k
- setting up and solving an equation to find doubling time or half-life using natural logarithms
- calculating future values or determining when a threshold value is breached
- Identifying a long-term limitation of an unconstrained exponential model, for example that P(t) = P₀e^(kt) predicts P → ∞ as t → ∞.
- Explaining why the model becomes unrealistic, such as finite resources limiting growth or ambient temperature setting a floor for cooling.
- Proposing a specific mathematical refinement, such as a carrying capacity K in a bounded growth model or a baseline shift θ_room in a decay model.
- Interpreting the refined model's parameters, for example stating that K is the limiting population or that θ_room is the ambient temperature.
- Comparing the original and refined predictions over a stated time horizon to justify why the refinement is needed.
Examiner Tips
- 💡Draw the curve getting progressively closer to the x-axis without ever touching or crossing it.
- 💡Label the y-intercept (0, 1) clearly on every sketch of an untransformed exponential function, and mark the asymptote y = 0.
- 💡Remember that the derivative of eˣ is simply eˣ, with no extra coefficient.
- 💡Label the coordinates (0, 1) and mark the asymptote y = 0 explicitly on sketches.
- 💡Remember: to differentiate e^(kx), multiply by the derivative of the power k.
- 💡When explaining suitability, state: 'The rate of increase or decrease is proportional to the amount present'.
- 💡To convert a logarithm to index form: start with the base, raise it to the other side, equate to the argument, e.g. y = logₐ x becomes aʸ = x.
- 💡Check all solutions to logarithmic equations to reject values that make any argument non-positive, and state the domain explicitly.
- 💡Remember that ln 1 = 0 and ln e = 1; label these reference points on sketches.
- 💡State the vertical asymptote equation x = 0 clearly when sketching y = ln x.
- 💡Isolate the exponential or logarithmic term completely before applying the inverse operation.
- 💡Remember: taking ln of both sides requires applying ln to the entire side, not term-by-term.
- 💡Before combining logs with addition or subtraction, move any numerical coefficients inside as powers.
- 💡Make sure all logarithms have the same base before applying the laws of logarithms.
- 💡Combine logs with addition into a product of arguments before exponentiating.
- 💡Always substitute final x-values back into original arguments to confirm they are strictly positive.
- 💡Clear all multiplying coefficients first using the power law: write 2 ln x as ln(x²).
- 💡Remember that −ln x = ln(1/x) and (1/2) ln x = ln(√x).
- 💡Keep logarithms in exact symbolic form (ln 2, ln 5) throughout your working until the final step.
- 💡Use brackets around binomial powers: write (x + 1) ln 2 so you remember to expand both terms.
- 💡Check the axes of the given graph: if both axes have logs, it is y = axⁿ; if only the y-axis has logs, it is y = kbˣ.
- 💡Calculate gradient using points far apart on the line of best fit for maximum accuracy.
- 💡Substitute t = 0 first: e⁰ = 1, which immediately reveals the initial value y₀.
- 💡Use exact values like ln 2 when finding doubling time or half-life to preserve accuracy.
- 💡Consider the behaviour of the model as t → ∞ to identify where predictions diverge from reality.
- 💡Refer to physical factors such as carrying capacity or ambient temperature when evaluating limitations, and state the amended model explicitly.
Common Mistakes
- Allowing the curve to cross the x-axis or curl away from the horizontal asymptote y = 0; the curve approaches the axis asymptotically.
- Drawing exponential decay curves with a vertical asymptote instead of approaching y = 0 as x → +∞.
- Forgetting that the y-intercept is always (0, 1) for any base a > 0, since a⁰ = 1.
- assuming e is an algebraic variable rather than Euler's constant e ≈ 2.71828
- drawing y = eˣ as touching or crossing the x-axis for negative values of x
- differentiating e^(kx) using the power rule, incorrectly writing kx e^(kx−1)
- forgetting the factor of k from the chain rule, writing d/dx(e^(3x)) = e^(3x)
- Attempting to evaluate the logarithm of zero or a negative number, which is undefined in real numbers; correction: always require the argument x > 0 before solving.
- Confusing the base and the argument when rewriting in index form, e.g. writing log₂ 8 = 3 as 8ʸ = 2; correction: the base of the log becomes the base of the power, so 2³ = 8.
- Treating the specification's x ≥ 0 as the domain; correction: x = 0 must be excluded because no real power of a gives 0, so the domain is x > 0.
- drawing the graph of ln x crossing into the negative x half-plane
- assuming ln x has a horizontal asymptote as x → ∞, when it actually grows without bound
- applying ln to each term individually in a sum, writing ln(A + B) = ln A + ln B
- taking exponentials across an equation incorrectly, writing e^(A + B) = e^A + e^B instead of e^A e^B
- assuming log(x + y) equals log x + log y, which is a false distributive assumption
- forgetting that the subtraction law gives a fraction inside a single log, not a division of two logs
- writing log x + log y = log(x + y) instead of log(xy)
- retaining negative roots that make an original logarithmic argument negative or zero
- writing log x − log y = (log x) / (log y) instead of log(x/y)
- attempting to combine terms with subtraction before moving outer coefficients into powers
- dropping brackets when expanding powers with logarithms, writing x + 1 ln 2 instead of (x + 1) ln 2
- evaluating logarithms prematurely and introducing decimal rounding errors into intermediate steps
- confusing power models (plot ln y against ln x) with exponential models (plot ln y against x)
- forgetting to exponentiate the intercept, setting a = c instead of a = eᶜ or a = 10ᶜ
- using a positive rate constant k in a decay context where the exponent should be negative
- forgetting to state units (such as years, hours, or grams) in final contextual answers
- Giving vague everyday criticisms without addressing the mathematical behaviour of the function as t → ∞; correction: state the predicted limit, e.g. P(t) → ∞, and explain why that is impossible.
- Stating that the model fails without proposing a constructive refinement; correction: offer a specific amended model, such as a bounded growth curve or a baseline-shifted decay curve.
- Confusing the roles of the parameters in a refined model; correction: identify K as the carrying capacity and θ_room as the ambient temperature before interpreting predictions.