The Derivative of ArcSecant or Inverse Secant is used in deriving a function that involves the inverse form of the trigonometric function ‘*secant*‘. **The derivative of the inverse secant function is equal to 1/(|x|√(x ^{2}-1)).** We can prove this derivative using the Pythagorean theorem and algebra.

In this article, we will discuss how to derive the arcsecant or the inverse secant function. We will cover brief fundamentals, its definition, formula, a graph comparison of the underived and derived function, a proof, methods to derive, and a few examples.

##### CALCULUS

**Relevant for**…

Learning about the proof and graphs of the derivative of arcsec of x.

##### CALCULUS

**Relevant for**…

Learning about the proof and graphs of the derivative of arcsec of x.

## Avoid confusion in using the denotations arcsec(x), sec^{-1}(x),
1
/
sec(x) , and sec^{n}(x)

It is important that we do not compromise the potential confusions we might have in using different denotations between $latex \text{arcsec}(x)$, $latex \sec^{-1}{(x)}$, $latex \frac{1}{\sec{(x)}}$, and $latex \sec^{n}{(x)}$, since interchanging the meaning of these symbols may lead to derivation mistakes. Summarizing the definition of these symbols, we have

$latex \text{arcsec}(x) = \sec^{-1}{(x)}$

Both symbols $latex \text{arcsec}$ and $latex \sec^{-1}$ can be used interchangeably when computing for the inverse secant of either a variable or another function. $latex \text{arcsecant}$ is commonly used as the verbal symbol of inverse secant function which is popularly used as introductory denotations for beginners whilst $latex \sec^{-1}$ is used as a mathematical symbol of inverse secant function for a more professional setting.

However, when it comes to the denotation $latex \sec^{-1}{(x)}$, sometimes it can confuse learners that $latex -1$ is an algebraic exponent of a non-inverse secant, which is not true. The $latex -1$ used for inverse secant represents the secant being inverse and not raised to $latex -1$. This has been proven and shown in the previous sub-article written above.

Therefore,

$latex \sec^{-1}{(x)} \neq \frac{1}{\sec{(x)}}$

And givens such as $latex \sec^{2}{(x)}$ or $latex \sec^{n}{(x)}$, where *n* is any algebraic exponent of a non-inverse secant, **MUST NOT** use the inverse secant formula since in these givens, both the 2 and any exponent *n* are treated as algebraic exponents of a non-inverse secant.

## Proof of the Derivative of the Inverse Secant Function

In this proof, we will mainly use the concepts of a right triangle, the Pythagorean theorem, the trigonometric function of secant and tangent, and some basic algebra. Just like in the previous figure as a reference sample for a given right triangle, suppose we have that same triangle $latex \Delta ABC$, but this time, let’s change the variables for an easier illustration.

where for every *one-unit* of a side adjacent to *angle y*, there is a side $latex \sqrt{x^2-1}$ opposite to *angle y* and a hypothenuse *x.*

Using these components of a right-triangle, we can find the *angle y* by using Cho-Sha-Cao, particularly the secant function by using the hypothenuse *x* and its adjacent side.

$latex \sec{(\theta)} = \frac{hyp}{adj}$

$latex \sec{(y)} = \frac{x}{1}$

$latex \sec{(y)} = x$

Now, we can implicitly derive this equation by using the derivative of trigonometric function of secant for the left-hand side and power rule for the right-hand side. Doing so, we have

$latex \frac{d}{dx} (\sec{(y)}) = \frac{d}{dx} (x)$

$latex \frac{d}{dx} (\sec{(y)}) = 1$

$latex \frac{dy}{dx} (\sec{(y)}\tan{(y)}) = 1$

$latex \frac{dy}{dx} = \frac{1}{\sec{(y)}\tan{(y)}}$

Getting the tangent of *angle y* from our given right-triangle, we have

$latex \tan{(y)} = \frac{opp}{adj}$

$latex \tan{(y)} = \frac{\sqrt{x^2-1}}{1}$

$latex \tan{(y)} = \sqrt{x^2-1}$

We can then substitute $latex \sec{(y)}$ and $latex \tan{(y)}$ to the implicit differentiation of $latex \sec{(y)} = x$

$latex \frac{dy}{dx} = \frac{1}{\sec{(y)}\tan{(y)}}$

$latex \frac{dy}{dx} = \frac{1}{(x) \cdot \left(\sqrt{x^2-1}\right)}$

$latex \frac{dy}{dx} = \frac{1}{x\sqrt{x^2-1}}$

Now, since

$latex \sec{(y)} = x$

and

$latex hypothenuse = x$

We know that a negative hypotenuse cannot exist. Therefore, $latex \sec{(y)}$ in this case cannot be negative. That’s why the x multiplicand in the denominator of the derivative of the inverse secant must be considered an absolute value.

$latex \frac{dy}{dx} = \frac{1}{|x|\sqrt{x^2-1}}$

Therefore, algebraically solving for the *angle y* and getting its derivative, we have

$latex \sec{(y)} = x$

$latex y = \frac{x}{\sec}$

$latex y = \sec^{-1}{(x)}$

$latex \frac{dy}{dx} = \frac{d}{dx} \left( \sec^{-1}{(x)} \right)$

$latex \frac{dy}{dx} = \frac{1}{|x|\sqrt{x^2-1}}$

which is now the derivative formula for the inverse secant of *x*.

Now, for the derivative of an inverse secant of any function other than *x*, we may apply the derivative formula of inverse secant together with the chain rule formula. By doing so, we have

$latex \frac{dy}{dx} = \frac{d}{du} \sec^{-1}{(u)} \cdot \frac{d}{dx} (u)$

$latex \frac{dy}{dx} = \frac{1}{|u|\sqrt{u^2-1}} \cdot \frac{d}{dx} (u)$

where $latex u$ is any function other than *x*.

## How to derive an Inverse Secant Function?

The derivative process of an inverse secant function is very straightforward assuming you have already learned the concepts behind the usage of the inverse secant function and how we arrived to its derivative formula.

### METHOD 1: Derivative of the Inverse Secant of any single variable *x*

$latex \frac{d}{dx} \left( \sec^{-1}{(x)} \right) = \frac{1}{|x|\sqrt{x^2-1}}$ |

**Step 1:** Analyze if the inverse secant of a variable is a function of that same variable. For example, if the right-hand side of the equation is $latex \sec^{-1}{(x)}$, then check if it is a function of the same variable x or f(x).

* Note:* If $latex \sec^{-1}{(x)}$ is a function of a different variable such as

*f(t)*or

*f(y)*, it will use implicit differentiation which is out of the scope of this article.

** Step 2:** Illustrate derivation by using the derivation symbols or denotations such as

$latex \frac{d}{dx} f(x) = \frac{d}{dx} \left(\sec^{-1}{(x)} \right)$

or

$latex \frac{dy}{dx} = \frac{d}{dx} \left(\sec^{-1}{(x)} \right)$

** Step 3: **Then directly apply the derivative formula of inverse secant function

$latex \frac{dy}{dx} = \frac{1}{|x|\sqrt{(x^2-1)}}$

If nothing is to be simplified anymore, then that would be the final answer.

### METHOD 2: Derivative of the Inverse Secant of any function *u* in terms of *x.*

$latex \frac{d}{dx} \left( \sec^{-1}{(u)} \right) = \frac{1}{|u|\sqrt{u^2-1}} \cdot \frac{d}{dx} (u)$ |

** Step 1:** Express the inverse function as $latex F(x) = \sec^{-1}{(u)}$ or $latex F(x) = \text{arcsec}(u)$, where $latex u$ represents any function other than

*x*.

** Step 2:** Illustrate derivation by using the derivation symbols or denotations such as

$latex \frac{d}{dx} F(x) = \frac{d}{dx} \left(\sec^{-1}{(u)} \right)$

or

$latex \frac{dy}{dx} = \frac{d}{dx} \left(\sec^{-1}{(u)} \right)$

** Step 3:** Consider $latex \sec^{-1}{(u)}$ as the outside function $latex f(u)$ and $latex u$ as the inner function $latex g(x)$ of the composite function $latex F(x)$. Hence we have

$latex f(u) = \sec^{-1}{(u)}$

and

$latex g(x) = u$

** Step 4:** Get the derivative of the outer function $latex f(u)$, which must use the derivative of inverse secant $latex \sec^{-1}{(u)}$, in terms of $latex u$.

$latex \frac{d}{du} \left( \sec^{-1}{(u)} \right) = \frac{1}{|u|\sqrt{u^2-1}}$

** Step 5:** Get the derivative of the inner function $latex g(x) = u$. Use the appropriate derivative rule that applies to $latex u$.

** Step 6:** Apply the basic chain rule formula by algebraically multiplying the derivative of outer function $latex f(u)$ by the derivative of inner function $latex g(x)$

$latex \frac{dy}{dx} = \frac{d}{du} (f(u)) \cdot \frac{d}{dx} (g(x))$

$latex \frac{dy}{dx} = \frac{1}{|u|\sqrt{u^2-1}} \cdot \frac{d}{dx} (u)$

** Step 7:** Substitute $latex u$ into $latex f'(u)$

** Step 8:** Simplify and apply any function law whenever applicable to finalize the answer.

## Graph of Inverse Secant *x* VS. The Derivative of Inverse Secant *x*

Given the function

$latex f(x) = \sec^{-1}{(x)}$

the graph is illustrated as

And as we know by now, by deriving $latex f(x) = \sec^{-1}{(x)}$, we get

$latex f'(x) = \frac{1}{|x|\sqrt{x^2-1}}$

which is illustrated graphically as

Illustrating both graphs in one, we have

Analyzing the differences of these functions through these graphs, you can observe that the original function $latex f(x) = \sec^{-1}{(x)}$ has a domain of

$latex (-\infty,-1] \cup [1,\infty )$ or *all real numbers except *$latex -1 < x < 1$

and exists within the range of

$latex [0,\frac{\pi}{2}\big) \cup \big(\frac{\pi}{2},\pi]$ or $latex 0 \leq y \leq \pi$ *except* $latex \frac{\pi}{2}$

whereas the derivative $latex f'(x) = \frac{1}{|x|\sqrt{x^2-1}}$ has a domain of

$latex (-\infty,-1) \cup (1,\infty)$ or *all real numbers except *$latex -1 \leq x \leq 1$

and exists within the range of

$latex (0,\infty)$ or $latex y > 0$

## Examples

Below are some examples of using either the first or second method in deriving an inverse secant function.

### EXAMPLE 1

**Derive:** $latex f(\theta) = \sec^{-1}{(\theta)}$

**Solution:** Analyzing the given inverse secant function, it is only an inverse secant of a single variable $latex \theta$ raised to a variable equal to one. Therefore, we can use the first method to derive this problem.

**Step 1:** Analyze if the inverse secant of $latex \theta$ is a function of $latex \theta$. In this problem, it is. Hence, proceed to step 2.

**Step 2:** Illustrate derivation through

$latex \frac{d}{d\theta} f(\theta) = \frac{d}{d\theta} \left(\sec^{-1}{(\theta)} \right)$

**Step 3: **Directly apply the derivative formula of inverse secant function and derive in terms of $latex \theta$. Since no further simplification is needed, **the final answer is:**

$latex f'(\theta) = \frac{1}{\theta\sqrt{\theta^2-1}}$

### EXAMPLE 2

**Derive:** $latex F(x) = \sec^{-1}{\left(8x^2-4 \right)}$

**Solution:** Analyzing the given inverse secant function, it is an inverse secant of a polynomial function. Therefore, we can use the second method to derive this problem.

**Step 1:** Express the inverse secant function as $latex F(x) = \sec^{-1}{(u)}$ or $latex F(x) = \arcsec{(u)}$, where $latex u$ represents any function other than *x*. In this problem,

$latex u = 8x^2-4$

We will substitute this later as we finalize the derivative of the problem.

**Step 2:** Illustrate derivation by using the derivation symbols or denotations such as

$latex \frac{d}{dx} F(x) = \frac{d}{dx} \left(\sec^{-1}{(u)} \right)$

**Step 3:** Consider $latex \sec^{-1}{(u)}$ as the outside function $latex f(u)$ and $latex u$ as the inner function $latex g(x)$ of the composite function $latex F(x)$. For this problem, we have

$latex f(u) = \sec^{-1}{(u)}$

and

$latex g(x) = u = 8x^2-4$

**Step 4:** Get the derivative of the outer function $latex f(u)$, which must use the derivative of inverse secant, in terms of $latex u$.

$latex \frac{d}{du} \left( \sec^{-1}{(u)} \right) = \frac{1}{|u|\sqrt{u^2-1}}$

**Step 5:** Get the derivative of the inner function $latex g(x)$ or $latex u$. Since our $latex u$ in this problem is a polynomial function, we will use power rule and sum/difference of derivatives to derive $latex u$.

$latex \frac{d}{dx}(g(x)) = \frac{d}{dx} \left(8x^2-4 \right)$

$latex \frac{d}{dx}(g(x)) = 16x$

**Step 6:** Apply the basic chain rule formula by algebraically multiplying the derivative of outer function $latex f(u)$ by the derivative of inner function $latex g(x)$

$latex \frac{dy}{dx} = \frac{d}{du} (f(u)) \cdot \frac{d}{dx} (g(x))$

$latex \frac{dy}{dx} = \frac{1}{|u|\sqrt{u^2-1}} \cdot 16x$

**Step 7:** Substitute $latex u$ into $latex f'(u)$

$latex \frac{dy}{dx} =\frac{1}{|u|\sqrt{u^2-1}} \cdot 16x$

$latex \frac{dy}{dx} = \frac{1}{|(8x^2-4)|\sqrt{(8x^2-4)^2-1}} \cdot 16x$

**Step 8:** Simplify and apply any function law whenever applicable to finalize the answer.

$latex F'(x) = \frac{1}{|(8x^2-4)|\sqrt{(8x^2-4)^2-1}} \cdot 16x$

$latex F'(x) = \frac{16x}{(8x^2-4)\sqrt{(8x^2-4)^2-1}}$

$latex F'(x) = \frac{16x}{4(2x^2-1)\sqrt{(8x^2-4)^2-1}}$

$latex F'(x) = \frac{4x}{(2x^2-1)\sqrt{(8x^2-1)^2-1}}$

And **the final answer is:**

$latex F'(x) = \frac{4x}{(2x^2-1)\sqrt{(8x^2-4)^2-1}}$

## See also

Interested in learning more about the derivatives of trigonometric functions? Take a look at these pages:

- Derivative of arcsin (Inverse Sine) With Proof and Graphs
- Derivative of arccos (Inverse Cosine) With Proof and Graphs
- Derivative of arctan (Inverse Tangent) With Proof and Graphs
- Derivative of arccsc (Inverse Cosecant) With Proof and Graphs
- Derivative of arccot (Inverse Cotangent) With Proof and Graphs

### Learn mathematics with our additional resources in different topics

**LEARN MORE**