Calibration Methods

By Stephanie Myers

Table of Contents
Introduction
Solution Preparation
Directly from a primary standard
Dilution from a stock
Serial Dilution
Dilution of Samples
Standard Calibration Curves (External Standard Curve)
Standard Addition
Mathematical
Graphical
Internal Standard Curve
With external standards
Constant concentration of internal standard
Variable concentration of internal standard
With standard addition
How do you choose?
Number of samples
Complexity of matrix
Instrument response
Summary

Practice problems
Solutions to practice problems (separate file)

Appendices
A. Glossary of terms
B. Graphing with Excel
Calculating precision of line with Excel

1
Introduction

Analytical chemistry is the branch of chemistry that seeks to answer the questions: “What is it?”

(qualitative analysis) and “How much of it is present?” (quantitative analysis). The substance these

questions are asked about, the “it”, is called the analyte. In instrumental analysis, detection is based on

the response of an instrument, the signal, to some property of the analyte. This distinguishes it from

classical analysis which is based on a chemical reaction of the analyte with other substances. In classical

analysis, the amount of analyte can be determined based on stoichiometric (mole‐to‐mole)

relationships. However, in instrumental analysis the relationship between signal and analyte does not

have a convenient theoretical relationship. In addition, it is often affected by the analyte’s matrix, the

substances, including the solvent, that surround the analyte.

Therefore, the effectively use instrumental analysis, the relationship between the analyte and

the signal must be established. The process of doing so is called calibration. While calibration can be

used to verify or establish signal response that is related to the identity of the analyte, it is more

typically used to determine the relationship between signal and analyte concentration. This primer

concentrates on methods to effectively determine that quantitative relationship between signal and

amount of analyte.

Determining the relationship of signal to analyte concentration requires a series of solutions

with a known concentration, standards. The signal of each standard can then be measured to

determine the relationship between signal and concentration. Once established, the relationship can

then be used to determine the concentration of analyte in a sample, a substance where the

concentration of analyte is unknown.

2
Solution Preparation

Because instrumental analysis requires a large number of solutions, it is critically important to

keep track of how each solution was made. Since the solutions generally look alike, it is useful to get

each solution a designation when you make it. Then reference to the solution (as it is connected with

the signal) is easier. One method of keeping track of solutions is a “solution box.” The solution box has

the name (designation of the solution) on the top, a list of components that include experimental

measurements with precision (significant figures) and total volume. Below the box, the concentration of

analyte can be noted. Figure 1 shows an example of the solution box. The calculations for the

concentration of that solution are shown in Example 1.

For a large number of solutions where each solution is made in a similar manner, a table may be

more convenient. Even with a table, it is still important to have a column for solution names. If each

entry in a column is the same, a note of that consistent quantity can be shown in at the top of bottom of

the table. Examples of using tables to describe solutions are shown in Examples 2 and 3.

Preparation from primary standard. Ultimately, each standard should start with a primary

standard. A primary standard is pure substance containing analyte that can be used to make solutions.

One method of preparing a series of standards would be to quantitatively dissolve various amounts of

primary standard. However, this method is not commonly used because the very low concentrations

typically needed for instrumental analysis are difficult to produce with reasonable precision. Instead

this method is often used to produce a stock solution. A stock solution is a standard with a

3
concentration that is higher than the standards that will actually be used in the experiment. Stock

solutions are used in the other two solution preparation methods.

Typical units for instrumental analysis are parts per million (ppm) which is the grams of analyte

in a million grams of solution. This is used for small (trace) components of a sample. If major

components of the sample are to be measured, percent (grams of analyte in one hundred grams of

solution) is an appropriate unit. For very small concentrations, parts per billion (ppb) or parts per trillion

(ppt) can be used. Molarity is rarely used in instrumental analysis as mole‐to‐mole relationships are not

relevant and it is not a typical reporting unit for official reports (which are rarely reviewed by chemists).

Example 1. Preparation of a stock solution.

If 0.0451 g of ferrous ammonium sulfate hexahydrate (FAS) was dissolved in a volumetric flask
to make 250.0 mL of solution, what is the concentration of iron in the stock solution in parts
per million (ppm) and molarity?

1 1
0.0451 ∙6 0.000115
392.13 1
0.000115
0.000459
0.2500
55.85 1000
0.000115 6.41
1 1
10
6.41
25.6
0.2500

As seen in Example 1, the relationship between analyte and primary standard must be

considered. If using a salt as your primary standard, it may be one of the ion concentrations that is

relevant to your experiment rather than the salt itself. If so, you must make sure you are calculating the

right concentration (that of ion rather than of the entire salt)! Calculations throughout the experiment

will be easier if the stock solution is described with the units used in the ultimate experiment.

4
 However, for a series of standards that are of reasonably high concentration and narrow range

of concentrations, dilution of a primary standard can be a useful method of standard preparation. The

major advantage of this method is that each solution is independent of the others. This is, if an error is

made in during the creation of one solution, the other solutions may still be correct. Example 2 shows a

typical set of data for creation of stock solutions from a primary standard.

Example 2. Standards from primary standard.

Table 1. Chloride ion solutions prepared from primary standard

Solution Mass of NaCl (g) [Cl‐] (M) [Cl‐] (ppm)
number
1 0.0567 0.00970 344
2 0.1011 0.01730 613.3
3 0.1488 0.02546 902.6
4 0.2699 0.04618 1637
5 0.3331 0.05700 2021
All solutions were prepared in 100.0 mL volumetric flasks

Sample calculation for solution 1

1 1 35.45 1000
0.0567 0.000970 34.4
58.44 1 1 1
0.000970 34.4
0.00970 344
0.1000 0.01000

Dilution from a stock. This is the most common method of solution preparation in instrumental

analysis. While the concentration range is still narrow, this does allow solutions to be made in low

concentrations with reasonable precision and reasonable ease. However, any error in the stock solution

will be transferred to all your standards. Because this error is so consistent across your standards, it is

normally undetectable. Thus the preparation of the stock solution itself is a critical step.

When using dilution from a stock, most calculations are made using the dilution equation

C1V1 = C1V2 (1)

5
Where C = concentration and V = volume for solutions 1 and 2 (the concentrated and the dilute).

Although chemists often prefer to write the equation with “M” instead of “C”, any concentration unit

can be used provided the same unit is used for both solutions 1 and 2. Similarly, for “V” any volume unit

can be used. In fact, if the solution component (bottom of the fraction) of concentration is a mass, you

can even use mass units instead of volume units in this equation. However, the equation should only be

used for dilution, it cannot be used in stoichiometric calculations (like titrations).

Example 3. Dilution from a stock.

Table 2. Standards prepared from iron stock solution

For stock solution [Fe] = 25.6 ppm
Total volume of each solution = 50.00 mL
solution Volume of Fe [Fe] (ppm)
stock (mL)
A 1.00 0.512
B 2.00 1.02
C 5.00 2.56
D 10.00 5.12

Sample calculation for solution B

C1V1 = C2V2
(25.6 ppm)(2.00 mL) = C2(50.00 mL)
1.02 ppm = C2

Serial Dilution. Although often confused with dilution from a stock, serial dilution is when each

diluted sample is used to make the next sample. This is most easily illustrated with solution boxes as

shown in Figure 2. Serial dilution is generally used when you wish to use a large concentration range.

Although not illustrated in Figure 2, it is frequently used when concentrations need to cover several

orders of magnitude (factors of 10). A major concern with serial dilution is that an error in one solution

will affect the concentration of all subsequent solutions.

6
 Preparation of the sample. It is rare that a sample is in an appropriate form for direct analysis.

Most instruments measure liquid solutions and the sample may be in a solid form or too concentrated

for analysis by the instrument or both. Therefore, it is critical to record in detail (solution boxes work

well!) any treatment done to the sample before measurement. The ultimate goal of the analysis is

determination of the analyte concentration in the original sample NOT the solution that was measured!

Normally, this is just the use of the dilution equation, but in reverse, finding the concentration

of the original solution not the diluted solution. If the original sample was a solid, it may be necessary to

multiply by the volume of solution to obtain the mass of analyte in the sample. A typical series of

calculations is illustrated in Example 4.

Forgetting to calculate the concentration in the original sample or only partly doing the

calculations is the most common error in instrumental analysis.

7
Example 4. Determination of analyte concentration in original sample.

A solution was made by dissolving 0.1234 g of an impure sample containing iron with
10.00 mL of HCl and diluting to a total volume of 100.0 mL. Then 5.00 mL of this solution
and 7.00 mL of various reagents were diluted to a total volume of 50.00 mL. The diluted
solution was measured by spectroscopy and determined to have a concentration of 2.68
ppm Fe. What is the concentration of iron in the sample?

Solution
First clarify the descriptions using solution boxes:

Determine concentration of Fe in solution 1

C1V1 = C2V2
C1(5.00 mL) = (2.68 ppm)(50.00 mL)
C1 = 26.8 ppm = 26.8 mg/L
Determine mass of Fe in solution 1
(26.8 mg/L)(0.100 L) = 2.68 mg x (1 g/1000 mg) = 0.00268 g
Determine the percent of Fe in sample
.
100 2.17%
.

8
Standard Curves

Standard curves also called calibration curves and external standard curves are the traditional

form of calibration. The creation of these curves requires a series of external standards for which the

signal of each external standard has been measured.

An external standard is a solution that contains a known concentration of the analyte. It is

separate (external) from the sample. Ideally, the matrix of the external standard should match as

closely as possible the matrix of the samples to be measured as the matrix can affect the measurement.

The calibration curve is a plot of the signal from the instrumental measurement (on the y‐axis)

versus the concentration of analyte (x‐axis). Signal is often, although not always, unitless. The units of

concentration should be specified and, of course, constant. Before using a calibration curve (any graph

in fact), it is important to look at it.

While most calibration curves are linear, this cannot be assumed. While the correlation

coefficient (R2) is a statistical measure of linearity, values of R2 deemed acceptable when the deviation is

due to random error may actually be due to the data curving instead. LOOK AT THE GRAPH to

determine if your data is linear or if it curves. If your data indicates a curve, there are two options: 1)

non‐linear curve fitting can be used to analyze your data or 2) (more common) the data can be

manipulated to create a line. Typical data manipulations are to graph the reciprocal of the data (1/x or

1/y) or to take the log of the data (log x or log y). In fact, one of the more familiar calibration curves, a

Beer’s Law Plot, actually does this. The instrument’s signal is actually transmittance (T) and the log of

this value, absorbance (A = ‐log T), is the quantity that is graphed. Because this is the normal use of the

spectrometer (instrument), the instrument does the calculation so analyst is sometimes unaware of this

manipulation.

Also, the range of concentrations where the signal‐concentration relationship is linear is

normally limited. At high concentrations, the detector reaches its maximum output and is said to be

9
“flooded.” Thus at sufficiently high concentrations, the line will flatten as the instrument gives the same

signal regardless of concentration. At low concentrations, the signal may become indistinguishable from

the noise (random error of the instrument signal). This can also cause the line to appear flat.

Consequently, the range in which the calibration shows a linear response extends from the limit of

quantitation (LOQ, lowest value of concentration that can be reliably related to a signal) to the limit of

linearity (LOL, the highest value of concentration that can be reliably related to signal). Limit of

quantitation is often confused with the limit of detection (LOD). At the limit of detection, analyte can be

detected as present, but a value cannot be reliably assigned to its concentration, just that the

concentration is above the LOD value.

In addition to inspecting the graph for linearity, it should also be evaluated for outliers. Outliers

are data points that do not follow the trend of the graph. These are easily identified visually but very

difficult to find with statistics. Normally, outliers are due to some experimental error and should be

discarded from the data analysis. However, this is a practice that should be used sparingly. It is too easy

to choose data points to fit a preconceived idea of what the data should look like. Discarding too many

points makes the data as unreliable as discarding none, regardless of the statistical values calculated. It

is better to repeat the experiment than to discard most of its data.

While statistical measurements are rarely useful to see linearity or find outliers, once you do

have a linear set of data, linear regression is the best tool to determine the equation of the line and the

associated precision. Linear regression finds the line that minimizes the distance between each data

point and the proposed line. It then summarizes these differences. In Excel, this calculation can be

performed with the “LINEST” function (see Appendix for directions on how to use this function). This

function will return the slope (m) and y‐intercept (b) and the errors associated with each (em and eb,

respectively). These errors are the best measure of the precision of your line. (Correlation coefficient,

R2, does NOT measure precision!) Propagation of error can be used to translate the error from the line

10
to the error for your answer. Assuming that the signal (y) was only measured once or has a negligible

error, the error in concentration (ex) can be determined using the equation:

(2)

In addition, it is normally a good assumption that the dilutions and other measurements that were part

of the sample preparation have errors much smaller than the equation of the line. In that case, the final

answer can be used in place of “x” in equation 2 and the result (ex) will be the error of your final answer.

180.00

160.00

140.00

120.00

100.00
intensity

80.00

60.00

40.00

20.00

0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
concentration (ppm)

Figure 3. Calibration curve for emission measurement of iron. Equation of the line is:
I = (4.44+0.23)C + (2.25+4.52)

11
 Example 5. Determination of concentration from a standard curve.
A solution was created by dissolving 0.1883 g of sample in water to make 100.0 mL of solution.
This solution was measured spectroscopically and produced an emission of 95.50. What is the
concentration of iron in the original sample, with error?

First calculate the answer:

Using the equation of the line:
I = 4.44C + 2.25
95.50 = 4.44C + 2.25
93.25 = 4.44C
21.0 ppm = C
Take into account solution prep
(21.0 mg/L)(0.1000L) = 2.10 mg Fe x (1 g/1000 mg) = 0.00210 g
(0.00210 g/0.1883 g) x 100 = 1.11%
Then calculate error:

. .
√0.002683 0.0023495 0.7094
. . . .
e = (0.0709)(1.11%) = 0.078
Final answer = 1.11±0.08%

A typical calibration curve is shown in Figure 3 and an example calculation for the concentration

of concentration using the calibration curve is shown in Example 5.

Standard Addition

One of the difficulties of making a standard curve is that composition of the external standards

need to match the composition of the sample as closely as possible. Nonanalyte components of the

sample that can affect the signal are called interferences. Interferences can affect the signal in various

ways. For example, substances can react with the analyte and reduce the signal. Alternately, the

interference can produce a signal itself increasing the overall signal. Such effects are normally

accounted for as part of the calibration curve. However, if little is known about the composition of the

sample or if the sample is particularly complex, creating external standards that account for

interferences is difficult or impossible.

12
 These problems can be overcome by the use of standard addition. With this method of

calibration, the same amount of sample is in each solution that is measured, therefore the same amount

of matrix is in each sample. Variations in the amount of analyte can be created by spiking the sample,

that is adding variable amounts of analyte to the sample. By adding analyte while keeping the amount

of sample constant, the matrix stays the same for each measurement.

Mathematical methods of standard addition. One method of standard addition is to directly

compare the measurement of the sample alone to the signal of a spiked sample. This method makes a

couple of assumptions: 1) that signal (S) is proportional to analyte concentration (c), or in math terms,

S = kc and 2) that such a small amount of analyte is added that the matrix is not significantly changed.

To be sure the first condition is met, it helps to know something about your sample and the

method to be sure that the concentrations are in the appropriate range. Also, the signal of your blank, a

solution with the same matrix but no analyte, must actually be zero. This actually rarely occurs, but can

be easily compensated for by using a corrected signal, that is the signal of the sample minus the signal of

the blank. Although the equations may not always specifically say so, in mathematical standard

addition, the signal used is always the corrected signal.

In a typical standard addition experiment, the signal of the sample (S1) is measured then a small

(measured!) amount (Vs) of concentrated standard (with concentration = cs) is added to a measured

amount of sample (Vx). The signal of the mixture is then measured (S2). In both measurements, the

signal is proportional to concentration with that proportionality constant (k). The two corrected signal

measurements can be compared as in equation 3.

(3)

The proportionality constants, k, will cancel. Obviously, c1 is the concentration of analyte in the

sample. However, the concentration of analyte in the spiked solution, solution 2 (c2), is a combination of

the analyte in the sample (c1) and the analyte in the standard (cs). However, concentrations cannot be

13
added directly. Therefore c2 must be expressed in terms of how the solution was made so that either

mass or moles of analyte is what is being summed. If the spiked solution was made in the manner

described above, its concentration can be determined from:

(4)

This equation does assume that volumes are additive. This is not normally true, however it is not a bad

assumption if the solutions are very similar (same solvent) and a small volume of standard is added. The

concentration of analyte in the sample (c1) can be determined by putting equations 3 and 4 together.

Example 6. Mathematical standard addition.

The signal when the sample was measured was 0.345. When 5.00 mL of sample was mixed
with 0.10 mL of 67.0 ppm standard, the signal was 0.891. What is the concentration of analyte in the
sample?

Putting equations 3 and 4 together:

. . .
. . .
0.891 0.345
. .

0.891c = 0.06764(5c + 6.70) 0.891c = 0.3382c + 0.4531 0.553c = 0.4531

c = 0.819 ppm

Another variation of this method is to dilute both the sample and the spiked sample to the same

volume (VT). In this way, the assumption that volumes are additive is not necessary and the math gets a

little easier. Also, since the same amount of sample is added to both solutions, the matrix of the

solutions is exactly the same. Thus the only assumption we are making is that signal is proportional to

concentration. If the experiment is set up this way, the equation becomes:

(5)

14
 Example 7. Mathematical standard addition, version 2
When 5.00 mL of sample was diluted to a total volume of 10.00 mL, the signal was 0.213. When 5.00
mL of sample was spiked with 0.10 mL of 67.0 ppm analyte and diluted to 10.00 mL, the signal was 0.452.
What is the concentration of analyte in the sample?

. . . . .
. .

2.26c = 1.065c + 1.4271 1.195c = 1.4271

c = 1.19 ppm

Graphical standard addition. There are two major issues with standard addition. The first is

the assumption that concentration is proportional to signal. (The assumption might not be true.) The

second is that it is difficult to determine the error associated with the measurement. While propagation

of error can be used, it does not always capture all of the associated errors. Both of these issues can be

addressed using graphical methods. The linear relationship can be confirmed by visualization of the line

and regression analysis can measure the error associated with the method.

Graphical standard addition requires a series of standards with the same total volume (VT), the

same amount of sample (Vx) and variable amounts of analyte standard (Vs). The signal is then graphed

15
against the concentration of added analyte. Preparation of solutions and the resulting graph are shown

in Example 8.

Example 8. Graphical standard addition. Solutions and graph.

Table 3. Solutions for standard addition. The following solutions were prepared in 50.00 mL
volumetric flasks and a 105.6 ppm Cu2+ standard was used.
solution Volume Volume Added [Cu2+] Signal
sample standard (ppm) intensity
(mL) (mL)
A 10.00 0.00 0.00 101123
B 10.00 1.00 2.11 204567
C 10.00 2.00 4.22 302229
D 10.00 3.00 6.34 409001

Sample calculation for solution C

C1V1 = C2V2
(105.6 ppm)(2.00 mL) = C2(50.00 mL)
4.22 ppm = C2
450000

400000

350000

300000

250000
intensity

200000

150000

100000

50000

0
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00
concentration of added copper (II) ion (ppm)

Figure 4. Standard additon curve for copper (II) ion. Equation of the line is:
I = (48334+527)C + (101130+2082)

16
 To determine the concentration of analyte in the solutions tested, the negative of the x‐

intercept of the line is calculated. However, this is NOT the concentration of analyte in the sample. The

sample was diluted during solution prep. Assuming the original sample was used to prepare solutions,

all that is required is to account for this dilution. However, if your sample solutions was prepared from

original sample, you must account for that as well.

Example 9. Determination of sample concentration from standard addition curve.

A sample solution was prepared by dissolving 0.0187 g of copper ore in 10 mL of nitric acid and
diluting to a total volume of 100.0 mL. The sample was then analyzed as described in Example 8. What
is the percent of copper in the ore, with error?

From Example 8, the equation of the line from the standard addition graph was:
I = (48334±527)C + 101130±2082
To determine the x‐intercept, set y = 0
0 = 48334C + 101130
‐2.09 = C
Obviously, concentration is not negative,
The concentration of Cu in the diluted sample = 2.09 ppm
Account for dilution, 10.00 mL of sample was added to each 50.00 mL solution (see Example 8)
C1V1 = C2V2
C1(10.00 mL) = (2.09 ppm)(50.00 mL)
C1 = 10.5 ppm = concentration of copper in sample solution
To get it back to the ore
10.5 mg/L x (0.1000 L) = 1.05 mg x (1 g/1000 mg) = 0.00105 g
(0.00105 g/0.0187) g x 100 = 5.59%
To determine error, use the error from the equation of the line and assume the error of dilution was
negligible
e
x

527 2082
5.59 48334 0 101130
0.0233 ex = 0.13
.
Answer = 5.59±0.13%

17
Internal Standards

For many experiments, instrumental response is dependent on factors that the analyst has little

or no control over. This might be an inconsistent instrument response caused by manual injection or just

a detector that is very sensitive to environmental factors. Another example is a complex sample

preparation where transfer of all the sample from reagent solution to reagent solution is not always

complete. In addition, it is often useful to have verification of the instrument response. In any of these

circumstances, it is often convenient to use an internal standard.

An internal standard is a substance, added in a known amount, that produces a signal that can

be measured in the same experiment as the analyte, but is distinctly different. In spectroscopy, this

would mean measurement at a different, nonoverlapping wavelength. In chromatography, the internal

standard would have a different retention time than the analyte. The assumption is that whatever

happens to the signal of the analyte will also happen to the signal of the internal standard. Therefore, it

is generally a good idea to choose an internal standard that is chemically similar to your analyte.

Constant amount of internal standard. Normally, the internal standard is added in a constant

amount. This means that its signal should also be constant. So when the calibration curve is

constructed, the ratio of the signal of the analyte divided by the signal of the internal standard is

graphed on the y‐axis. This way, if the signals are reduced (for whatever reason), the lower internal

standard signal will bring up the analyte signal to its proper value. Higher signals would be compensated

for in the same manner. Also, because the internal standard concentration is constant, the ratio of

signals depends only on the concentration of analyte. Therefore, you can add an internal standard to

either an external standard curve or a standard addition curve by simply graphing the ratio of signals

(signal of analyte/signal of internal standard) instead of the signal of analyte alone. The addition of

internal standard must be the same for sample solution as it was in all the external standards. If

18
complex solution preparation is the issue to be solved, the internal standard should be added before the

sample preparation begins.

Solution preparation for an internal standard experiment of this type looks like Table 4.

Example 10. Internal Standard Solution Preparation.

An analyte stock solution has a concentration of 189.6 ppm Cu. A 159.1 ppm of yttrium (Y) was
prepared for use as an internal standard since it was unlikely to be in the sample. Prepare a series of
standards for an external standard calibration curve that have a range of 5 – 50 ppm copper and use a
constant concentration of yttrium as an internal standard for an atomic emission experiment.

Solution

Table 4. Solutions for external standard/internal standard calibration curve. Each solution is made in a
50.00 mL volumetric flask.
Solution Volume of Cu Volume of Y [Cu] (ppm) [Y] (ppm)
stock (mL) stock (mL)
A 1.50 5.00 5.69 15.9
B 5.00 5.00 19.0 15.9
C 9.00 5.00 34.1 15.9
D 13.00 5.00 49.3 15.9

How much stock should be used?

Low end: C1V1 = C2V2 (189.6 ppm)(V) = (5 ppm)(50.00 mL) V = 1.319 mL
Round, it doesn’t have to be exact as long as the value is known
High end: C1V1 = C2V2 (189.6 ppm)(V) = (50 ppm)(50.00 mL) V = 13.19 mL
Even it out with values in the middle and recalculate with actual values and the dilution
equation
Example for solution C: C1V1 = C2V2 (189.6 ppm)(9.00 mL) = (C)(50.00 mL) C = 34.1 ppm

19
Example 11. Solving with internal standard data.
1.811 g of a sample containing copper was dissolved using nitric acid and diluted with water
to a total volume of 200.0 mL. 5.00 mL of 159.1 ppm Y standard was mixed with 15.00 mL of this
sample solution and the mixture diluted to a total volume of 50.00 mL. Data as shown in Table 5
was collected for the standards created in Example 10 and this sample solution. What is the
concentration of copper in the original standard?

Solution
Table 6. Emission Data for Copper with Yttrium as an internal standard. Total volume of each
solution is 50.00 mL.
Volume Volume
intensity intensity
of Cu of Y [Cu] [Y] ratio
Solution at 324 at 190
stock stock (ppm) (ppm) ICu/IY
nm (Cu) nm (Y)
(mL) (mL)
A 1.5 5 5.69 15.9 1.401 0.922 1.52
B 5 5 19 15.9 1.978 0.968 2.043
C 9 5 34.1 15.9 2.230 0.866 2.575
D 13 5 49.3 15.9 2.998 0.938 3.196
sample
15 5 0.973
2 2.662

Solution
First graph the results and determine the equation of the line as shown below.
3.500

3.000
intensity at 324 nm/intensity at 190 nm

2.500

2.000

1.500

1.000
0.0 10.0 20.0 30.0 40.0 50.0 60.0
concentration of copper (ppm)

Figure 5. Emission spectroscopy of copper with yttrium as an internal standard. The equation
of the line is: I = (0.0381+0.0007)C + (1.303+0.024)

20
 Example 11 continued. Calculations based on graph of results.
For the solution labeled “sample 2” the ratio of the signal is: 2.662/0.973 = 2.736.
This is the “I” used in the equation of the line: I = 0.03810C + 1.303
So 2.736 = 0.03810C + 1.303
1.433 = 0.03810C
37.61 = C
Then you have to account for dilution from the original sample solution:
C1V1 = C2V2 C1(15.00 mL) = (37.61 ppm)(50.00 mL) C1 = 125.4 ppm
In the 200.0 mL of original sample solution there is
125.4 mg/L x (0.2000 L) = 25.07 mg x (1 g/1000 mg) = 0.02501 g Cu
So in the original solid sample there is:
0.02501 g/1.811 g x 100 = 1.384% Cu

Variable amounts of internal standard. Sometimes it is not convenient to add the same

amount of internal standard to each solution. This is typical when (for whatever reason) you are adding

internal standard with a mass instead of a volume measurement. To compensate for this variation,

simply graph the ratio of concentration of analyte to concentration of internal standard on the x‐axis.

The only addition is multiplying “x” when you solve the equation of the line by the concentration of

internal standard so that you are working with the correct value for your other calculations.

21
Example 12. Internal standard calculations with variable internal standard.
A series of managanese standards used iridium as an internal standards. Atomic emission
measured the intensity of the standard manganese (IMn) and standard iridium (IIr) for each
concentration. When the results were graphed, the following equation of the line was determined:
0.697 0.008 0.055 0.031

A 1.776 g sample and 0.078 g iridium were dissolved in nitric acid and diluted to 100.0 mL. A 5.00
mL aliquiot of this solution was diluted to 50.00 mL and measured. The intensity at the manganese
wavelength was 7.211 and the intensity at the iridium wavelength was 7.606. What is the percent
manganese in the sample?

Solution
.
0.948
.
Using the equation of the line
0.948 = 0.697X + 0.055
0.893 = 0.697X
1.28 = X
But this is for the diluted solution
Use the dilution equation to get the ratio of concentrations in the original solution (since both
internal standard and sample were diluted)
C1V1 = C2V2 C1(5.00 mL) = (1.28)(50.00mL) C2 = 12.8
Since in the original solution, 0.078 g Ir was diluted to 100.0 mL, this means the concentration of
iridium was: 78 mg/0.10 L = 780 ppm
12.8 CMn = 9984 ppm
9984 mg/L x (0.10L) = 998.4 mg x (1 g/1000 mg) = 0.998 g Mn
To get percent Mn
0.998 g Mn/1.776 g sample x 100 = 56.2%
Error
. .
0.01862 ex = (0.01864)(56.2) = 1.046
. . .

Final answer = 56.2±1.0%

How do you know to use ppm? Look at the units on the slope!

22
 Internal Standard with Standard addition. The internal standard measurements and standard

addition experiments can be combined. In this experiment, solutions are made with constant amounts

of sample and of internal standard and each solution is spiked with a different concentration of analyte.

The signal for the analyte and internal standard are measured separately. The calibration curve is a

graph with the ratio of the analyte signal to the internal standard signal (y‐axis) versus the concentration

of added analyte.

Example 13. Internal standard with standard addition

A 3.351 mg sample was dissolved to make 100.0 mL of aqueous solution. A 126 ppm Cr
stock solution and a 359 ppm Sm stock solution were used to make the following 25.00 mL solutions
for analyis:
Table 7. Standard Addition/Internal standard analysis of chromium
solution volume sample volume Cr volume Sm concentration of
(mL) stock (mL) stock (mL) added Cr (ppm)
A 5.00 0.00 3.00 0
B 5.00 0.100 3.00 0.50
C 5.00 0.200 3.00 1.00
D 5.00 0.300 3.00 1.51

When the absorbance values of these solutions were measured and the results graphed, the
resulting equation of the line was:
1.983 0.016 1.708 0.008
What is the concentration of chromium in the sample?

Solution
0 = 1.983C + 1.708
‐0.8613 = C
Accounting for the dilution where there was 5.00 mL of sample in each 25.00 mL solution
C1V1 = C2V2 C1(5.00 mL) = (0.8613 ppm)(25.00 mL) C1 = 4.31 ppm
So in the original sample solution there were
4.31 mg/L x (0.1000L) = 0.431 mg Cr
In the solid
0.431 mg Cr/3.351 mg sample x 100 = 12.9%
error
. .
0.00933 ex = (0.00933)(12.9) = 0.120
. . .
Final answer = 12.9±0.1%

23
How to Choose the Best Calibration Method

There are three primary criteria to consider when choosing the best calibration method. Often

more than one of these criteria will be important. In that case, it is up to the analyst to decide which

criteria takes precedence, or to combine calibration methods to account for both criteria. Occasionally,

none of these issues makes a significant impact on the analysis. In that case, any calibration method is

acceptable.

Number of samples. Despite the usual academic experience, real analyses typically require that

the method is going to be used repeatedly for a large number of samples. For example, in a

manufacturing environment, samples may be collected every hour, all day (24 hours), all week (7 days)

in order to monitor the manufacturing process. Many experiments require the collection of a large

amount of data in order to fully understand the system under study.

When there are a large number of samples to be analyzed, an external standard calibration

curve is generally the method of choice. The advantage of this method is that the same number of

standards (generally about 5) can be used for any number of samples. Therefore, the time consuming

process of creating solutions can be minimized. Table 8 compares the relationship between the number

of solutions (standards and samples) and number of samples to be analyzed for external standard and

standard addition calibration, assuming that 5 data points are used in a line.

Table 8. Number of solutions needed for calibration methods.

number of external standard standard
samples standard addition addition
(math) (graph)
1 6 2 5
2 7 4 10
5 10 10 25
10 15 20 50
20 25 40 100
100 105 200 500

24
 In addition, the standards can be reused as calibration checks periodically during

measurements. In other words, after the calibration curve has been established, it is good practice to

include a standard or two in each batch of samples. The signal of the standard can be checked against

the calibration curve to be sure that the concentration‐signal relationship has not changed.

Complexity of Matrix. Standard addition is the technique designed to account for the

complexity of the matrix. Because the same amount of sample is in each solution that is measured, the

matrix is consistent for each measurement.

A complex matrix is often due to the many components in the sample solution, normally in

significant concentrations. This often affects the viscosity of the solution, ionic strength, and acid‐base

interactions (particularly Lewis acid‐base interactions) in such a way that signal can be affected. To

recreate these solutions so that the external standards are representative of sample can be time‐

consuming, or (if the sample is not well‐characterized) impossible.

Instrument Response. For many instruments, signal is affected by factors beyond the analysts’

control. This is the situation designed for internal standards. Internal standards account for variation

in instrument response, consistency of analyte concentration in a complex sample preparation and

validate both the quantitative (signal) and qualitative (e.g., wavelength, retention time) values.

Summary

All instrumental methods require some form of calibration. There are several options for the

type of calibration to be used. The choice of calibration method depends on the analytical problem. All

require the preparation of solutions of known concentration (standards) and normally the sample itself

will also need to be modified before analysis. The method of solution preparation and data analysis will

depend on the type of calibration used. Ultimately, your analysis is only as good as the calibration,

therefore proper attention to this part of the analysis is critical to the quality of the results.

25
Practice Problems for calibration methods

Concentration Calculations
1. What is the concentration, in ppm copper, of the following solutions?
a. 0.05811 g copper (II) sulfate tetrahydrate in 100.0 mL of aqueous solution?
b. 0.0112 g copper metal dissolved in 10.0 mL of nitric acid and diluted to a total volume of
500.0 mL
c. 1.00 mL of 67.8 ppm copper stock solution diluted with water to 250.0 mL

2. Which method would be best to make a series of standard sodium ion solutions that have a
concentration ranging from 1 M to 1 x 10‐5 M ? Why?

3. Given a 237.2 ppm stock solution of hydroquinone, fill in the table below to make a series of
standards using the dilution form a stock method. Total volume of each sample is 50.00 mL
Table 3a. Dilution of hydroquinone.
Solution Goal concentration (ppm) Volume stock Actual
(mL) concentration
(ppm)
A 5.00
B 10.00
C 15.00
D 25.00

3b. Why does the actual concentration not equal the goal concentration? How close do you
need to get to the goal concentration? Justify your answer.

4. A 1.3246 g sample was dissolved with 5.00 mL of concentrated hydrochloric acid. The resulting
solution was diluted with water to a total volume of 100.0 mL. A 25.00 mL aliquot of this
solution was neutralized with sodium hydroxide and then 15.00 mL of 0.10 M ethylenediamine
was added. The mixture was diluted to a total volume of 200.0 mL. This mixture was analyzed
spectroscopically and determined to have a concentration of 1.56 ppm Co. What is the percent
of cobalt in the original sample?

26
5. Determine the concentration of potassium in ppm for each solution described by the solution
boxes below:

External Standard Calibration

6. An analyst performed an external standard calibration for nitrate and determined the following
line:
Signal = (7.230±0.0183ppb‐1) (concentration) + 0.7±0.8

a. If a sample produced a signal of 156.7, what is the concentration of nitrate in the

measured solution, with error?
b. If the sample measured had been diluted by a factor of 10 before measurement, what is
the concentration, with error, for the original sample?

7. An analyst was developing a method and so measured a large number of external standards to
determine if the signal to concentration relationship was linear, and in what range was it linear.
The data for this analysis is given below. Graph the data, determine which points can be validly
used for analysis then use the valid points to determine the equation of the line (with error!).

Table 7a. Analysis of carbonate ion.

Solution [CO32‐] Signal
(ppm)
1 0.00 13.99
2 1.01 14.03
3 3.03 15.55
4 5.07 18.03
5 10.10 35.71
6 15.05 54.03
7 20.03 70.04
8 24.07 84.56
9 29.03 105.70
10 33.06 113.19
11 45.01 160.11
12 59.05 210.67
13 75.07 263.03
14 91.08 270.01
15 100.01 275.54

27
8. A 5.378 g solid sample containing calcium chloride was dissolved in water to make 150.0 mL of
solution 1. Solution 2 was made by diluting 3.00 mL of solution 1 with water to make a total
volume of 50.00 mL. Solution 3 was made by mixing 2.00 mL of solution 2 with 15 mL of
reagent mixture and diluting to 25.00 mL. When solution 3 was measured spectroscopically for
chloride ion, the signal was 4.097.
The equation for the external standard curve:
Signal = 0.03651±0.00031ppm‐1(concentration) = 1.234±0.1031
What is the concentration of chloride in the original solid sample, with error?
Assuming all the chloride came from calcium chloride, what is the percent calcium chloride in
the original, solid, sample, with error?

Standard Addition
9. Write a formula to express the concentration of manganese ([Mn]) in a solution made by mixing
5.00 mL of a sample containing manganese that has a concentration of cx and 3.00 mL of a
standard manganese solution with a concentration of 15.00 ppm diluted to a total volume of
25.00 mL.

10. When an aqueous sample was analyzed for iron the signal was 5.994. When 0.200 mL of a 75.00
ppm iron standard was added to 10.00 mL of sample, the signal was 7.559. What is the
concentration of iron in the original sample?

11. 5.00 mL of a sample containing chromate ion was diluted to 25.00 mL and its absorbance at 383
nm was 0.132. When 5.00 mL of the same (original) sample was mixed with 3.00 mL of 100.3
ppm chromate standard and diluted to 25.00 mL, the resulting solution had an absorbance of
0.855. What is the concentration of chromate in the sample?

12. When analyzed for methanol, a sample had a signal of 0.9262. When 0.0134 g of pure methanol
was mixed with 1.7393 g of sample, the signal was 3.8677. What is the concentration of
methanol in the sample?

13. The following series of solutions were made. What is the concentration of cobalt in the sample,
with error?

Table 13a. Standard addition analysis of cobalt using a 43.8

ppm cobalt standard and a total volume of 50.00 mL.
Solution Volume Volume signal
sample standard
(mL) (mL)
1 10.00 2.00 10.22
2 10.00 4.00 15.37
3 10.00 6.00 19.37
4 10.00 8.00 26.43
5 10.00 10.00 30.78

28
 14. A series of solutions made with 20.00 mL of sample and spiked with various amounts of 36.71
ppm nickel(II) ion before dilution to a total volume of 100.0 mL and analyzed for nickel. The
signal was graphed against the concentration of added nickel(II) ion. The graph produced the
following equation:
Signal = (0.296±0.051 ppm‐1)[Ni2+] + 9.264±0.423
What is the concentration of nickel(II) in the sample (with error)?

Internal Standard
15. If you were doing an analysis for fluoride ion, which of the following would make the best
internal standard and WHY? State the advantages and disadvantages of each choice.
Chloride ion, bromide ion, phosphate ion, sodium ion, liquid bromine

16. Use the following description to graph and external standard calibration curve and determine
the concentration of zirconium(II) ion with error.
Table 16a. Analysis of Zr2+ using Hf2+ as an internal standard.
Standard concentrations were: [Zr2+] = 32.83 ppm; [Hf2+] = 37.23 ppm
and total volume of each solution was 50.00 mL
Solution Volume Volume Volume Signal Zr Signal Hf
Zr2+ std Hf2+ std sample
(mL) (mL) (mL)
1 3.00 4.00 0 0.1066 0.6046
2 5.00 4.00 0 0.2323 0.6928
3 7.50 4.00 0 0.3832 0.5987
4 10.00 4.00 0 0.5098 0.7001
X 0 4.00 15.00 0.4500 0.6995

17. A 20.00 g sample of ground mystery meat was extracted into 50.00 mL of ethanol. After
filtering, 10.00 mL of the ethanolic solution was diluted with water to make 25.00 mL of a
sample solution “X” that was analyzed as described in the table below. What is the
concentration of leucine in the mystery meat, with error?
Table 17a. Analysis of leucine using a synthetic amino acid (Z) as an
internal standard. The standard concentrations were: [leucine] = 198.3
ppm; [Z] = 229.1 ppm and the total volume of each solution was 100.0 mL.
Solution Volume Volume Volume Signal Signal Z
leucine of std Z of sample leucine
std (mL) (mL) X (mL)
S1 0.00 1.00 5.00 4336 3968
S2 0.50 1.00 5.00 5314 3859
S3 1.00 1.00 5.00 6480 4054
S4 1.50 1.00 5.00 7895 4189
S5 2.00 1.00 5.00 8108 3762

29
 18. A 15.00 g sample of ketchup was diluted to 250.0 mL. A standard addition experiment that used
10.00 mL of this diluted ketchup solution, 4.00 mL of 30.00 ppm oxalic acid and varying amounts
of a 45.00 ppm ascorbic acid standard with a total volume of 50.00 mL was analyzed and
resulted in a linear graph with a line equation of:
0.1808 0.0045 3.5235 0.0661
What is the concentration of ascorbic acid in the ketchup with error?

19. In the analysis of potassium, cesium ion was used as an internal standard. When 10.00 g of
banana was mashed up, 0.00100 g of cesium chloride was added to the banana. This mixture
was diluted to 250.0 mL and allowed to sit overnight to extract the potassium. After filtering,
100.0 mL of this mixture was evaporated down to 10.00 mL. The mixture was then analyzed
with ICP and an intensity of 536462 was measured for potassium and an intensity of 366965 was
measured for cesium. Using the calibration curve described below, determine the
concentration of potassium in the banana, with error. (Concentration units used were ppm.)
. . . .

Choosing the Correct Method

20. Consider the solution preps described below. Label each as: external standard, internal
standard or standard addition. For some examples, two of the choices may be correct.

Table 20a. Analysis of copper.

Soln [Cu] [Cr]
(ppm) (ppm)
1 10.0 5.00
2 20.0 5.00
3 30.0 5.00
4 40.0 5.00

Table 20b. Analysis of ethanol, using 100% ethanol standard; total volume 50.00 mL with water
as solvent.
Solution Volume Volume Volume
ethanol sample methanol
(mL) (mL) (mL)
1 1.00 25.00 10.00
2 2.00 25.00 10.00
3 3.00 25.00 10.00
4 4.00 25.00 10,00

30
Table 20c. Analysis of zinc using a 100.00 ppm zinc ion standard; total volume 25.00 mL.
Solution Volume
Zn (mL)
A 2.00
B 5.00
C 10.00
D 20.00

Table 20d. Analysis of silver using a 46.4 ppm standard solution of silver ion.
Solution Volume Volume Total
sample silver volume
(mL) standard (mL)
(mL)
A 15.00 0.50 20.00
B 15.00 1.00 20.00
C 15.00 1.50 20.00
D 15.00 2.00 20.00

21. Consider the problem scenarios described below. For each scenario, list whether each of the
three criteria for choosing a method is significant or unimportant (or somewhere between). Use
this information to justify your answer to: “Which calibration method would be best to solve it?”

A. A new method of quality control at a Kool‐Aid plant is being developed. The concentration
of blue dye in the powdered product will be measured spectroscopically.
B. There is a persistent theory that Napoleon died of arsenic poisoning, either on purposed
from his enemies or accidentally from the arsenic used as a pigment in his green wall paper.
Since heavy metals are retained in the hair, the body will be dug up and ICP‐OES will be used
to analyze his hair to determine the concentration of arsenic.
C. Augusta has an old water system. There is concern that there might be lead contamination
of the water.
D. An old painting was found in a barn. Perhaps it is one of the old masters? The pigments in
the paint will be tested to see if the orange has a concentration of iron consistent with the
appropriate time period.
E. An academic researcher was trying to determine the concentration of nitrate in surface
water and well water surrounding a farm. Unfortunately, the ion chromatograph available
only had manual injection, so that the sample size of each measurement would be
inconsistent.

31
Appendix A. Glossary

Analyte = substance of interest that is being measured

Blank = solution that does not contain analyte, but ideally has the same matrix as the sample

Calibration = process of determining the relationship between signal and a property (normally amount)
of an analyte

Calibration curve = a graph of signal (y‐axis) versus concentration (x‐axis) produced by measuring the
signal of a series of external standards; same as standard curve or external standard curve

Classical analysis = analysis based on traditional chemistry, generally some type of reaction

Corrected signal = signal minus signal of the blank

Correlation coefficient (R2) = a statistical measure of linearity; the closer the value is to one, the more
closely the set of data corresponds to a line

External standard = a solution with a known concentration of analyte; does not contain any sample

External standard curve = a graph of signal (y‐axis) versus concentration (x‐axis) produced by measuring
the signal of a series of external standards; same as calibration curve or standard curve

Flooding the detector = producing a signal so high that the detector cannot distinguish between
differences in concentration

Instrumental analysis = analysis based on an instrument’s response to some property of an analyte

Interference = a substance found in the sample which is not the analyte but still affects signal

Internal standard = substance, added in a known amount, that produces a signal that can be measured
in the same experiment as the analyte, but is distinctly different

Limit of detection = the lowest concentration where the method can reliably determine the presence
(although not necessarily the concentration) of analyte

Limit of linearity = the highest concentration that can reliably be measured by the method

Limit of quantitation = the lowest concentration value that can be reliably measured by the method

Linear regression = see regression analysis

Matrix = substances that surround the analyte. This includes the solvent, other reagents and any
contaminants

Molarity = moles of solute per liter of solution

32
Major component= a substance that occurs in large (normally mid‐percent) concentrations in the
sample

Noise = instrument response that is not related to the analyte

Order of magnitude = factor of ten; the “a” when scientific notation is expressed as: N x 10a

Outlier = data point that does not follow the trend of the rest of the data

Parts per billion (ppb) = parts in one billion parts. Officially ppm = (g solute/g solution) x 109 which
through convenient math is the same as μg/kg or ng/g. Since water has a density where 1 kg = 1L, in
dilute aqueous solution the formula ppm = μg/L or ng/mL is often used

Parts per million (ppm) = parts in one million parts. Officially ppm = (g solute/g solution) x 106 which
through convenient math is the same as mg/kg or μg/g. Since water has a density where 1 kg = 1L, in
dilute aqueous solution the formula ppm = mg/L or μg/mL is often used

Primary standard = a pure substance, containing analyte, that can be used to make standard solutions

Qualitative analysis = experiments designed to determine the identity of an analyte

Quantitative analysis = experiments designed to determine the concentration of an analyte

Range = the concentration values where the signal has a proportional (linear) relationship to
concentration; limit of quantitation to the limit of linearity

Regression analysis = a statistical means of determining the equation of a line that minimizes the
distance of the line and each data point, the difference between the data and the line is summarized as
an error in the slope and y‐intercept

Sample = a substance that is to be analyzed where the amount of analyte is unknown

Secondary standard = standard solution where the concentration was determined either by another
experiment or by creating a solution from something other than a primary standard

Serial dilution = preparation of a series of solutions where each diluted solution is used to make the
next dilution; normally used when concentrations need to over a wide (orders of magnitude)
concentration range

Signal = response of an instrument to an analyte

Spike = a known amount of analyte added to a sample

Standard = a solution with a known concentration of analyte

Standard addition = calibration method where a known amount of analyte is added to the sample

33
Standard curve = a graph of signal (y‐axis) versus concentration (x‐axis) produced by measuring the
signal of a series of external standards; same as calibration curve or external standard curve

Stock solution = a standard solution of high concentration that will be used to make other standard
solutions

Trace component = substances in the sample that occur in small (usually low ppm) concentrations

34
Appendix B. Using Excel to graph a line

Graph Criteria:
Use spacing effectively: The graph size should be more than ½ page.
Only show quadrants actually used or needed for context
Title/Caption:
Includes appropriate numbering
Labeled as “Figure” not “graph” or whatever
Does not repeat axes
(Ask instructor if you do not know the name of the graph type)
Include the identity of what you are measuring
Put title (more appropriately called a figure caption) at bottom of table
Figure caption should not overlap other components of graph
Some common graph types: calibration curve, Beer’s Law Plot, Spectrum
Axes: The x‐axis is used for the independent variable
The y‐axis is used for the dependent variable
Do not include axes or axes sections where it is not used to graph data
(or show context)
Labeled (label is not unit, but includes unit)
Include dimension (unit) in parenthesis
Have visible tick marks
Spacing between tick marks reasonable for reading
Normally 1, 2 or 5 divisions
Tick labels have correct significant figures (consistent decimal places)
Data points: Visible but not inappropriately large
Data points reflect actual data
All the data!
Include outliers on graph even if they were ignored when drawing trend line.
If more than one data set is on the same graph, use different marker styles
Trend: Show data trend using line or curve as appropriate
If more than one data set is used,
use a different style line for each and include a legend.
Do NOT include a legend if only one data set is used.
If trend is linear, determine equation of the line

35
Creating the graph described above, with an example

Step 1. Input your data

Typically “X” data is in column A and “Y” data is in column B
You may use the first row for a heading

Input the following data:

Table 1. Electrochemical Calibration of PO43‐

trial [PO43‐] (M) potential (mV)
1 0.1000 67.5
2 0.0500 55.1
3 0.0100 38.1
4 0.0050 30.0
5 0.0010 18.0

To add superscripts or subscripts…

Go to the “home” tab
In the lower right corner of the font section there is a little box with an arrow
Click on the little box to bring up a pop up
In the pop‐up click on the appropriate box, then “ok”
Now you add the superscript or subscript
Uncheck the box to stop using a super or subscript
Step 2. Make a graph
a. Highlight the entire data set
(numbers not column headings; trial number is not part of your data)
b. Choose “insert” tab
c. Under “insert charts” choose the scatter chart (Other name: “xy”)
It’s the icon that looks like a bunch of dots.
WARNING! Do not choose “line graph” it is wrong and will not allow you to fix it.
Clicking on the arrow will give you a bunch of scatter plot choices
Normally choose the one with dots only (first choice)
For complex graphs (like titration curves)
choose the one with the dots and curve through them.
d. The chart will pop up on the spreadsheet.
This is normally inconvenient and hard to see so let’s move it.
Choose “move chart” from the chart tools/design menu bar
Click “New sheet”
You may want to name it…how about “Figure 1”?

Step 3. Make it pretty

a. Add gridlines
Click on the plus sign on the right side of the chart
Click on the little triangle/arrow to the right of “gridlines”

36
 Check “major gridlines” for both horizontal and vertical axes
Optional: You can also check minor gridlines if you do so for BOTH axes

b. Fix axes
i. Insert the axes titles
Click on the plus sign next to the chart
Click to check “axes titles”
A box should appear at each axis
Click in the box and type in the appropriate title
This title should exactly match the column heading of your data table
To add superscripts or subscripts…
Go to the “home” tab
In the lower right corner of the font section there is a little box with an arrow
Click to check the box and it will let you add the superscript or subscript
Uncheck the box to stop using a super or subscript

ii. Make sure the axes have the correct number of significant figures
Double click on one of the axes (or choose “more options” from the axes list of the plus sign)
A “format axis” box should pop up on the right hand side
Choose “axes options” and the bar chart (at top right under title)
Choose the “number” category (almost at bottom, scroll down)
Go to “category” then choose “number”
Input the number of decimal places appropriate to your measuring device/data
(if your values are in scientific notation, choose “scientific”)

iii. OPTIONAL under the “format axes” menu

You can add “tick marks” (not required if you have gridlines, essential if you do not)
You can change the spacing of your gridlines under “units”
(best choices are divisions of 1, 2 or 5)
You can change the maximum and minimum values under “bounds”

iv. Repeat parts ii and iii above for the other axis

v. IF YOUR AXIS IS IN THE WRONG SPOT

click on double‐click on the other axis
that’s right, the axis you don’t want to move—so to move the y‐axis
click on “format x‐axis”
choose “axis options” (bar chart as above)
Toward the end of that section, you will see “vertical axis crosses at”
Choose “axis value” and input the leftmost number from your x‐axis
when moving the x‐axis, format the y‐axis and tell the horizontal axis to cross at your lowest number.

37
c. Figure Caption
Click on the plus sign next to the chart
Click to check the box “chart title”
Fill in the appropriate information.
It should look a lot like your Table title but it is a “Figure” not a table.

Hover your mouse near a corner to the move icon with four arrows.
Move the title to the bottom of the graph window and left justify it.
Move (possibly resize) other graph elements so that everything is readable.
You may also want to re size the graph title so it looks good to you.

Step 4. Add a Trendline

a. Look at your data. Is the trend a line or a curve or something else?
Hint: The data given is NOT a line.

b. click on the plus sign to the right of the chart

Click on the arrow that pops up when you hover over “trendline”
Do not click on the “trendline” box! I will insert the line you don’t want!
Click on “more options”
In the format trendline box that pops up,
click on the type of curve that looks most like your data
Click in the box that says “display equation on chart”
Move equation to a visible location.

Step 5. Print the graph

Go to the “file” tab then choose print

Using Excel to do calculations

We prefer linear graphs, so let’s practice using Excel to do calculations
To make this example graph linear, we need to graph the log of the concentration instead of
concentration. However, we still want to graph potential NOT the log of the potential.
In addition, it is easier to create two new adjacent columns with “x” on the left and “y” on the
right than to tell Excel where you have put your data. So do this:

Choose another column (probably on the same sheet)

If you are going to do a formula in Excel, it must start with an equal sign.
Note the cell location of your first concentration value (in this example it is A2)
Also note your first potential value (in this example it B2)
In another cell type “=log(a2)”
It does not matter whether or not you capitalize the A
Do not type the quotes, type what is inside them
Remember to use the reference of the cell you want to change not my example
In the next cell to the right type “=b2”
Here you are just moving the number without changing it.

38
 Copy and Paste these two cells so that you have the same number of data points.
If you have the talent you can also do a copy and drag, but I never figured that out
If you did too many cells it will show up as zeros. You can delete or ignore those.

FYI: To multiply in Excel, you use the * symbol; to divide use the /. You can write formulas with these
terms based on the principles shown above. For example, to use the dilution equation. If the amount of
25.00 ppm stock solution was in cell b3, and the total volume was 75.00 mL, to get the diluted
concentration, write the formula
=b3*25/75
Like other calculating devices, Excel doesn’t understand significant figures.

To determine the equation of a line, with error, in Excel

Highlight 6 blank cells in a 2 column by 3 row grid
Click in the function (fx) line and type
=linest(
Then highlight your y data and type a comma
Highlight your x data and type a comma
Finish typing with: true, true)
The overall function line should look something like: =LINEST(B3:B10,A3:A10,TRUE,TRUE)
Then for PC press CONTROL+SHIFT+ENTER
Or for Mac press “COMMAND+RETURN” m b
Your highlighted cells will get values that represent: um ub
R2 sy

On the ribbon at the top, go to “insert” then choose the text box icon at the top.
Draw the text box where you want to put in the equation and (using the numbers from before)
type:
y = (m±um)x + (b±ub)
Or you can put this information in the figure caption.

Sneaky trick: You can make a + by underlining a “+” sign! Otherwise you can find it in the
“insert→symbol” section.

39