Chemistry IA code: FJ203K

Investigation on the Chemical Kinetics of the Silver Mirror Reaction

1 Introduction

The silver mirror reaction between Tollen’s reagent (diamminesilver(I) hydroxide solution,
[Ag(N H3)2]OH) and a reducing sugar (such as glucose) is a well known experiment that is both
visually appealing and useful for teaching the concept of redox reactions. When teaching local
middle school students chemistry through this reaction as part of my CAS program, I realized that
it took much longer for the mirror to form in their unheated classroom compared to the heated
lab in our school. This led me to think if I could explain quantitatively the chemical kinetics of
this reaction, so that in the future people won’t think they failed the experiment and give up just
because it is happening slower than expected. This investigation is meaningful also because the
reaction is used for Tollen’s test for reducing sugars and my research can help find the optimum
temperature for the test.

1.1 Research question

How does temperature change (20.0°C, 25.0°C, 30.0°C, 35.0°C, 40.0°C) affect the reaction
rate of the reduction of diamminesilver(I) ions ([Ag(N H3)2]+aq.)) to silver metal (Ag(s.)) by glucose
(
(C6H12O6(aq.)), as measured by the change in reflectance of the silver mirror formed?

1.2 Background information

1.2.1 Silver mirror reaction
The silver mirror reaction is a redox reaction between solutions of the oxidizing agent, diammi-
nesilver(I) hydroxide, and the reducing agent, glucose, in an alkaline solution. The half equations
are:

−
(aq.) + e → Ag(s.) + 2N H3(aq.)
[Ag(N H3 )2 ]+ reduction
− −
C6 H12 O6(aq.) + H2 O(l.) → C6 H11 O7(aq.) + +
3H(aq.) + 2e oxidation
The overall ionic equation and overall molecular equation are:

−
(aq.) + C6 H12 O6(aq.) + H2 O(l.) → 2Ag(s.) + C6 H11 O7(aq.) + 4N H3(aq.) + 3H(aq.)
2[Ag(N H3 )2 ]+ +

2[Ag(N H3 )2 ]OH(aq.) + C6 H12 O6(aq.) → 2Ag(s.) + C6 H12 O7(aq.) + 4N H3(aq.) + H2 O(l.)

In the reaction, the aldehyde group of the glucose molecule, denoted RCHO below, is oxidised
to a carboxylic acid group in gluconic acid (C6 H12 O7 ), denoted RCOOH. A reaction mechanism
is proposed by Benet et al. (675):

RCHO(aq.) + H2 O(l.) ⇌ RCH(OH)2(aq.) Step 1

− −
RCH(OH)2(aq.) + OH(aq.) → RCH(OH)O(aq.) + H2 O(l.) Step 2
− ·
RCH(OH)O(aq.) (aq.) → RCH(OH)O(aq.) + Ag(s.) + 2N H3(aq.)
+ [Ag(N H3 )2 ]+ Step 3
·
(aq.) → RCOOH(aq.) + Ag(s.) + 2N H3(aq.) + H(aq.) Step 4
+ [Ag(N H3 )2 ]+ +
RCH(OH)O(aq.)

1
 They found that the rate determining step is Step 2. The rate law of this step is

r = k0 [RCH(OH)2 ][OH − ], (1)

where k0 is the rate constant.

Because Step 1 is a fast step in equilibrium, the equilibrium constant, K, can be expressed as

[RCH(OH)2 ]
K= .
[H2 O][RCHO]

Rearranging for [RCH(OH)2 ] and substituting it back into Equation 1,

r = k0 K[H2 O][RCHO][OH − ].

Because [H2 O] can be treated a constant for dilute solutions, k0 K[H2 O] can be substituted
entirely by a constant k, giving the overall rate law of the whole reaction:

r = k[RCHO][OH − ]. (2)

1.2.2 Reflectance and chemical kinetics

As the reaction progresses, the silver metal is produced as nanoparticles that form a colloid in
water. As the particles get larger, they cluster and precipitate onto the walls of the reaction vessel,
forming a silver mirror. Therefore, the glass walls of the vessel turn from transparent to reflective
due to the mirror, which changes the reflectance of the mirror (the percentage of incident light
that is reflected). In this investigation, reflectance is measured by placing a piece of red paper
next to the reaction vessel and taking a video of the reflection of the paper. The reflectance is
taken as the red value in the RGB color code of the reflection as recorded by the camera.
The time needed to reach 95% of the final reflectance, t, can be used to quantify the rate of
reaction. Assuming that the mirror will reach this reflectance when the amount of silver formed
is n, the average rate is equal to
n
r= .
t
Equating this with the rate law deduced before (Equation 2),

n 1 k[RCHO][OH − ]
= k[RCHO][OH − ], ⇒ = . (3)
t t n
The rate constant k can be related to temperature T by the Arrhenius equation, where Ea is
the activation energy of the reaction, R is the gas constant, and A is a constant specific to this
reaction called the pre-exponential factor (Zhang 147):
Ea
k = Ae− RT .
Substituting this into Equation 3,

1 Ea A[RCHO][OH − ]
= e− RT · .
t n
Taking the natural logarithm on both sides, we find that ln 1t should have a linear relationship
with T1 :

1 Ea 1 A[RCHO][OH − ]
ln =− + ln . (4)
t R T n

2
1.3 Hypothesis
The hypothesis of the experiment is that the natural logarithm of the inverse of the time taken
for the silver mirror to reach 95% reflectance has a linear relationship to the inverse of the reaction
temperature if the concentration of all reactants remain constant.

2 Methodology
2.1 Variables
Independent variable:

• Temperature of the reactants. The temperature is set to 20.0°C, 25.0°C, 30.0°C, 35.0°C
and 40.0°C by placing the reactants in a water bath. The reaction is carried out when the
actual temperature, measured with 0 − 100°C thermometers, reaches the designated values.

Dependent variable:

• Time taken for the reaction to finish. A video of the reaction is taken and the
reaction time is calculated as the time difference between when the reactants are mixed
and when the mirror reaches 95% of the final reflectance, as marked by the timestamps in
the video. The reflectance-time relationship is measured using the red-paper-and-camera
method mentioned before to determine when the reflectance reaches this value.

Controlled variables:

• Concentration of diamminesilver(I) ions and glucose in the corresponding

solutions. The concentration of reactants affects the rate of reaction. [Ag(N H3 )2 ]+ is
controlled at 0.100mol dm−3 and glucose at 0.111mol dm−3 when preparing solutions.

• pH of the solutions. As deduced from the reaction mechanism, the concentration of

OH − affects the rate of reaction. 1.00g of N aOH is dissolved in the silver solution for each
250cm3 batch before it is added to the volumetric flask. The glucose solution is kept neutral.

• Volume of solutions used in each trial. The volume ratio between the solutions affect
the final concentration of reactants in the reaction mixture. 15.0cm3 of diamminesilver(I)
solution and 5.00cm3 of glucose solution are used for each trial.

• Surface area of the reaction vessel. This affects the area that the silver precipitates
on and therefore the rate of change of the reflectance. The reactions are carried out in
20cm3 cylindrical glass bottles of diameter 27mm and height 58mm from the same batch
manufactured by Shanghai Loikaw Instrument Co., Ltd.

2.2 Chemicals and apparatus

Chemical Description Quantity

AgN O3 crystals 5g
N H3 solution 25 − 28% 20cm3
N aOH flakes 5g
glucose anhydrous powder 5g

Table 1: Chemicals used in investigation. Quantities are slightly more than least amount needed.

3
 Glassware Measuring instruments
Name Size Quantity Name Size Quantity
beaker 100cm3 3 measuring 3
25cm (±0.3) 2
volumetric 250cm3 (±0.1) 1 cylinder 3
10cm (±0.05) 1
flask 100cm3 (±0.08) 1 thermometer 0 − 100°C (±0.5) 2
glass bottle 20cm3 15 electronic
500g (±0.01) 1
glass rod 1 balance
Other
water bath red paper plain paper fume hood distilled water bottle droppers
phone weighing paper computer

Table 2: Apparatus used in investigation.

2.3 Procedure
The procedure of the experiment is as follows:

1. Make the 0.100mol dm−3 diamminesilver(I) hydroxide (molar mass 158.94g mol−1 ) solution
(solution A):

(a) Weigh 4.25g of AgN O3 (molar mass 169.87g mol−1 ) using an electronic balance, and
add it to a 100cm3 beaker. Dissolve the solid in approximately 50cm3 of distilled water.
(b) Inside a fume hood, measure 15.0cm3 of concentrated ammonia solution with a 25cm3
measuring cylinder and add slowly to the AgN O3 solution. Wash the cylinder with
around 5cm3 of distilled water and pour the water into the solution. A brown precipi-
tate should form and mix the solution with a glass rod until the precipitate disappears
completely. The reaction here is:
+
2Ag(aq.) + 2N H3(aq.) + H2 O(l.) → Ag2 O(s.) (brown) + 2N H4(aq.)
+
(5)
−
Ag2 O(s.) + 4N H3(aq.) + H2 O(l.) → 2[Ag(N H3 )2 ]+
(aq.) + 2OH(aq.) (6)

(c) Weigh 1.00g of N aOH using an electronic balance. Add it to a separate 100cm3 beaker
and add approximately 30cm3 of distilled water to dissolve it.
(d) Pour both solutions into a 250cm3 volumetric flask. Use distilled water to wash both
beakers at least 2 times with approximately 10cm3 of distilled water each time and
pour the water into the flask.
(e) Top up the volumetric flask with distilled water to the 250cm3 mark. Plug in the
cap and mix the solution well. This is a 0.100mol dm−3 solution of diamminesilver(I)
hydroxide (solution A).

2. Make the 0.111mol dm−3 glucose (molar mass 180.16g mol−1 ) solution (solution B):

(a) Weigh 2.00g of the glucose with an electronic balance. Add the glucose to a 100cm3
beaker and add 30cm3 of distilled water to dissolve it.
(b) Pour the solution into a 100cm3 volumetric flask. Wash the beaker as before.
(c) Top up the volumetric flask with distilled water to the 100cm3 mark. Plug in the cap
and mix the solution well. This is a 0.111mol dm−3 solution of glucose (solution B).

3. Place the vertical pieces of red and white paper, the camera, and the 20cm3 glass bottle
according to Figure 1, making sure that the reflection can be captured by the camera.

4
 Figure 1: Experimental
setup viewed from above.

4. Measure 15.0cm3 of solution A using a 25cm3 measuring cylinder and pour it into the 20cm3
glass bottle. Measure 5cm3 of solution B using a 10cm3 measuring cylinder and keep the
solution inside. Place both the glass bottle and the 10cm3 cylinder in a 20.0°C water bath.
Place two thermometers into the two containers to monitor the temperature.

5. When the temperature of both solutions reach 20°C, start the recording on the camera
(phone). Quickly remove both thermometers, and place the glass bottle in its original
position in the setup. Add solution B in the measuring cylinder to the glass bottle. Put on
the cap of the bottle, shake strongly, and return it to the original position. Stop recording
when the shiny mirror on the glass walls has stopped changing for the at least 20 seconds.

6. Using a video processor (OpenCV was used for this investigation), analyze the RGB color
code of a point on the reflection of the red paper in the mirror against time. Record the
time marked on the video when the addition of solution B to the glass bottle starts (t1 )
and ends (t2 ), and the time when the r (red) value of the reflection has reached 95% of the
terminal value at the end of the video (t′ ).

7. Repeat steps 3-6 for 2 more times.

8. Repeat steps 3-7 for temperatures 25.0°C, 30.0°C, 35.0°C and 40.0°C.

2.4 Precautions
Safety: Concentrated ammonia solution is very volatile, producing ammonia gas that causes
respiratory irritation. It is also a weak base, so direct contact causes severe skin burns or eye
damage, and disposal into the environment can harm aquatic life (Sigma-Aldrich, “Ammonium
Hydroxide Solution” 2). Sodium hydroxide is a strong base and causes skin burns or eye damage
(“Sodium Hydroxide” 2). Silver nitrate is an oxidizer and causes skin burns (“Silver Nitrate” 2).
To ensure safety, parts of the experiment involving concentrated ammonia solution are performed
in a fume hood and gloves, goggles and a lab coat are always worn.
Diamminesilver(I) hydroxide decomposes to silver nitride over time, which is a contact ex-
plosive (Breedlove and Softy 281). Therefore, the solutions are prepared just before usage and
unused solution is treated with excess glucose solution before disposal to reduce silver(I) ions to
silver metal.
Environmental: The two bases involved can impact the local water pH if directly poured
down the drain, and silver nitrate is toxic to aquatic life (Sigma-Aldrich, “Silver Nitrate” 2), so
all waste liquid is disposed in designated waste containers.
Ethical: To minimise the amount of chemicals used so that the experiment is more sustain-
able, the reactions are carried out in 20cm3 bottles instead of large beakers.

5
3 Analysis
3.1 Raw data
3.1.1 Qualitative data
During the experiment, the following qualitative observations were made. After the glucose
was mixed, the solution turned gradually from colorless to yellow, brown, and then black. Next,
a silver-coloured layer starts to form on the walls of the vessel and the reflection on the mirror
gradually becomes clearer and brighter until it does not change eventually.
This is shown by the data collected in the
video analyzer. Initially, the r value at the po-
sition of the reflection is very high, because it is
the r value of the white paper background be-
hind the colorless solution. Then, it decreases
quickly due to the formation of yellow sub-
stances in the solution which makes the solu-
tion darker until it reaches black, where the r
value reaches a minimum. After this, the mir-
ror starts forming, and the brightness of the red
reflection, as shown by the r value, increases to
a terminal value which signals the end of the
reaction.
Comparing this to theory, the yellow sub-
stance is silver nanoparticles existing as col-
loids and the black color is caused by the for- Figure 2: Sample data of the change of the r
mation of large nanoparticles. value against time for τ = 20°C.

3.1.2 Quantitative data

The raw quantitative data obtained in the experiment is presented below.

Time
Temperature
Trial Start adding glucose End adding glucose 95% reflectance reached
τ ± 0.5/°C t1 ± 0.1/s t2 ± 0.1/s t′ ± 0.1s
6.0 7.1 101.9
1 20.0 5.6 6.7 105.2
4.2 5.4 103.3
7.4 8.3 70.9
2 25.0 2.6 3.5 70.0
7.6 8.4 76.9
2.9 3.7 59.1
3 30.0 8.5 9.4 55.9
3.0 4.2 57.7
3.8 4.7 40.3
4 35.0 2.7 3.6 46.2
2.3 3.3 37.3
8.8 9.7 37.3
5 40.0 2.3 3.0 32.7
7.1 7.8 38.2

Table 3: Raw data table.

6
 The full r-time relationships measured from the video analyzer are plotted in both Figure 2
and the Appendix.

3.2 Data processing

The raw data is processed to find the logarithm of the inverse of reaction time and the inverse
of the reaction temperature in the Kelvin scale. A sample calculation for Trial 1 is given below.
First, for Trial 1 Run 1, the time of the addition of glucose is taken as the middle value of
the start and end, and its uncertainty is half the difference between the start and end plus the
original measurement uncertainty:
6.0s + 7.1s 7.1s − 6.0s
t0 = = 6.5s, δt0 = + 0.1s = 0.6s.
2 2
The time for the reaction to finish, t is the difference between t′ and t0 :

t = t′ − t0 = 101.9s − 6.5s = 95.4s, δt = 0.6s + 0.1s = 0.7s.

The similar process is carried out for Runs 2 and 3, giving t = 99.0 ± 0.8s and 98.5 ± 0.8s.
The average value for t is then calculated, with the uncertainty being half the difference between
maximum and minimum values plus the largest uncertainty of the three individual values:
95.4s + 99.0s + 98.5s 99.0s − 95.4s
t= = 98s, δt = + 0.8s = 3s.
3 2
Finally, the inverse of the value is calculated and its natural logarithm is taken:
1 1
ln = ln = −4.58.
t 98s
The uncertainty of the logarithm is half the difference between the results when the calculation
is performed on the upper and lower bounds of t within the uncertainty range:

1 ln 1 − ln t+δt
1
ln 1 − ln 98s+3s
1
δ ln = t−δt = 98s−3s = 0.03.
t 2 2
As for the temperature, it is first converted into the Kelvin scale:

T = τ + 273.15K = 20.0K + 273.15K = 293.2 ± 0.5K.

Its inverse is then calculated:

1 1 1 0.5K
= = 3.411 × 10−3 K −1 , δ = · 3.411 × 10−3 K −1 = 0.006 × 10−3 K −1 .
T 293.2K T 293.2K

3.3 Processed data

The processed data is presented below. The data is then fitted with a linear model using
Logger Pro and plotted with Pyplot in Figure 3.

Trial 1
T
/× 10−3 K −1 ln 1t
1 3.411 ±0.006 -4.58 ±0.03
2 3.354 ±0.006 -4.19 ±0.05
3 3.299 ±0.005 -4.0 ±0.1
4 3.245 ±0.005 -3.6 ±0.1
5 3.193 ±0.005 -3.39 ±0.07

Table 4: Processed data table.

7
 Figure 3: Plot of experimental data.

The linear fit has a R2 value of 0.995, which suggests that the hypothesis of the Arrhenius
equation does work for the silver mirror reaction: ln 1t decreases with T1 linearly. The line of best
fit is
1 1
ln = K + B,
t T
where A and B are
K = −5400 ± 700K,
B = 14 ± 2.
The uncertainties are calculated as half the difference between the values for the maximum
and minimum gradient lines.

3.4 Calculation of activation energy and pre-exponential factor

According to Equation 4, the gradient should be equal to
Ea
K=− .
R
Therefore, rearranging the terms,
Ea = −KR = − (−5400K) · 8.314JK −1 mol−1 = 45 ± 6 kJmol−1 ,
δK 700K
δEa % = × 100% = × 100% = 12.8%.
K −5400K

8
 Also, from Equation 4, the y-intercept B should be theoretically equal to
A[RCHO][OH − ]
B = ln ,
n
with A being the pre-exponential factor in the Arrhenius equation.
This gives
n
A= eB .
[RCHO][OH − ]
The concentration of glucose in solution B is 0.111mol dm−3 , and 5cm3 of solution B is mixed
with 15cm3 of solution A, so the concentration of glucose in the reaction mixture is

5cm3
[RCHO] = 0.111mol dm−3 · 3 3
= 0.028 ± 0.001 mol dm−3 .
5cm + 15cm
Similarly, since 1g of N aOH was used for 250cm3 of solution A,

1g 15cm3
[OH − ] = · = 0.075 ± 0.005 mol dm−3 .
39.997gmol−1 · 250cm3 5cm3 + 15cm3
Also, assuming that when the reflectance reaches 95% of the terminal value the reaction has
finished, n should be the stoichiometric amount of silver metal produced from the reactants. Since
silverdiammine(I) ions are added in excess, the total amount of silver produced is equal to twice
the amount of glucose added by the stoichiometric ratio,

n = 0.111mol dm−3 · 5cm3 · 2 = 0.00111 ± 0.00002 mol.

Therefore, the pre-exponential factor of the Arrhenius equation is

n 0.00111mol
A= eB = · e14 = (6.0 ± 1.7) × 105 dm6 mol−1 s−1 ,
−
[RCHO][OH ] 0.028mol dm · 0.075mol dm
−3 −3
 
δn δ[RCHO] δ[OH − ] δB
δA% = + + + × 100% = 29.0%.
n [RCHO] [OH − ] B

3.5 Comparison with scientific context

The data points in the graph fit very well with the linear model, and the best fit line passes
through all error bars with no outliers. The gradient is negative and the y-intercept is positive,
which is consistent with the hypothesis. Therefore, this suggests that the experimental data
supports the Arrhenius equation.
The calculated activation energy is compared to the accepted results from other chemists.
Zhu et al. investigated how the reaction time for the silver mirror to form varies under different
temperatures (92). However, instead of glucose, they used ethanal as the reducing agent. The
same calculations are performed on their data and a value for the activation energy is calculated
as Ea = 44kJmol−1 .

τ /°C 95 90 85 80 75 70
t/s 30 45 50 60 70 95

Table 5: Raw data for temperature and reaction time from Zhu et al. (92).

The percentage error of my experimental value is:

Ea (experimental) − Ea (accepted) 45kJmol−1 − 44kJmol−1

error% = = = 3.4%.
Ea (accepted) 44kJmol−1

9
4 Conclusion
From my experimental data, I can conclude that the relationship between the rate constant
and temperature for the silver mirror reaction between diamminesilver(I) ions and glucose under
alkaline conditions follows the Arrhenius equation. The natural logarithm of the reciprocal of
reaction time (which represents the rate of reaction) under fixed concentration of reactants has a
negative linear relationship with the reciprocal of temperature in the Kelvin scale.
The activation energy and pre-exponential factor calculated from the experiment are Ea = 45±
6kJmol−1 (with percentage error 3.4% from accepted value) and A = (6.0±1.7)×105 dm6 mol−1 s−1
respectively.

5 Evaluation
5.1 Reliability
The linear fit in the graph is almost perfect with a coefficient of correlation at 0.995, suggesting
that the linear models fits very well with the experimental data of ln 1t against T1 . None of the
data points are very far away from the line of best fit and the line passes through the error bars
of all five points, confirming that there are no outliers.

5.1.1 Precision
In the graph, all five points have small error bars compared to the data range, which shows the
high precision of the measurements. The processed data has relatively small percentage uncer-
tainties, which confirm this point. The gradient of the best fit line has a percentage uncertainty
of δK% = 12.3% and the y-intercept has δB% = 15.9%.
The high precision of the data is because a digital method was used to measure the reaction
time and a relatively precise thermometer was used to measure the temperature. It further shows
that random errors involved in the experiment did not have a huge impact on the data.

5.1.2 Accuracy
The calculated activation energy is compared to the accepted value to evaluate the accuracy
of the experiment. The percentage error of 3.4% is significantly smaller than the percentage
uncertainty of the value for activation energy calculated from the experimental data, 12.8%. This
suggests that the accuracy of the experiment is good within the range of measurement precision
for this experiment.
The percentage error can be explained by the choice of reducing agents. The reactive functional
group for both glucose and ethanal are the aldehyde group, therefore both reactions should follow
the reaction mechanism proposed before. However, the rate determining step (Step 2) is a reaction
between a 1,1-diol (known as a geminal diol) formed from the aldehyde with a hydroxide ion. The
diol acts as a Brønsted-Lowry acid and donates a proton. The glucose molecule has a carbon
chain of length 12, while ethanal has a carbon chain of length 2. The longer carbon chain has a
stronger electron releasing effect, which makes the anion formed less stable, therefore decreasing
the strength of the acid. Since the geminal diol formed from glucose is a weaker acid, it would be
less reactive for the rate determining step, therefore it would require a higher activation energy.
By considering this difference, it can be concluded that the accuracy of the experiment is quite
good.

10
5.2 Strengths and weaknesses
5.2.1 Successful methodologies
• I decided to keep the concentration of the reactants in the reaction mixture as a controlled
variable. This enables me to use the rate of reaction under each temperature directly as the
rate constant. This is further equal to the reciprocal of reaction time, and this reduced the
number of measurements I needed to do to obtain reliable data.

• I used the reflectance of the silver mirror to quantify the rate of reaction. It is impossible to
determine when the reaction finishes by observation, so it was necessary to find a method to
quantify reflectance. This method only requires simple materials like paper and a camera,
and provides a reliable way to quantitatively measure the reaction time.

5.2.2 Error sources and improvements

Random errors:

• The cleanness of the glass bottle interior affects how quickly the silver precipitate deposits
on the walls. The more dirty and unsmooth the surface is the harder it is for a shiny silver
mirror to form from the silver nanoparticles, which affects the reflectance of the mirror
formed. This causes a random error in the rate of change of reflectance of the silver mirror.

Methodology: The bottles used are from the same batch manufactured by the same company
to make sure the glass has the same chemical composition and smoothness. The bottles
were washed thoroughly before the reaction to remove any impurities from the glass. This
makes sure that the surface of the glass is of the same smoothness and would not interfere
with the deposition of the silver particles.

• The glucose solution is not added instantaneously, which makes it hard to determine when
exactly the reaction starts. This causes a random error in the measurement of reaction
time.
Methodology: The glucose solution is poured as quickly as possible, while making sure that
it doesn’t spill out. Also, the time of pouring is calculated into the uncertainty of reaction
time.
Improvement: Use a syringe instead of pouring the glucose solution, which can add the
solution very quickly. A syringe should be readily available in the lab. However, some of
the solution might be left in the syringe when adding it and introducing this additional
equipment leads to a systematic error in the volume of solution used.

Systematic errors:

• The reaction was carried out outside the water bath so that the reflectance can be measured
with the camera. The temperature of the reactants may decrease because of heat loss to the
surroundings, which leads to a systematic error in reaction temperature. This would cause
the actual reaction temperature to be less than the measured temperature of the solutions,
which was determined when the solutions were still in the water bath.
Methodology: The bottle with the diamminesilver(I) solution is moved out of the water
bath quickly and the glucose is added as soon as possible to reduce heat loss. It was made
sure that the camera and red paper were set up right next to the water bath, so that the
solutions do not need to be moved very far after leaving the water bath. Also, in order for
the reaction to be carried out as soon as possible, the camera setup is organized well in

11
 advance and the camera starts recording before the solutions are removed from the bath.
Improvement: In the future, it is possible to use a transparent water bath vessel so that
the reflectance can be measured while the bottle is in the bath. One experimental difficulty
that may arise from this is that the transparant vessel refracts and reflects the reflected
light from the silver mirror and it may be harder to capture a video of the reflection using a
camera outside the vessel. Alternatively, a waterproof camera may be used directly inside
the vessel.

• Diamminesilver(I) solution decomposes over time, which decreases the concentration of the
reactants. This will decrease the rate of reaction, especially for the trials that are carried
out later during the experiment.
Methodology: The diamminesilver(I) solution is made right before the experiment to min-
imise its decomposition. New solutions are always made for trials that are done on a separate
day. Also, during the experiment, the solution is stored in a capped volumetric flask to make
sure that the volatile solutes (dissolved N H3 gas) evaporate less quickly and the composition
of the solution remains as constant as possible.

• The solutions do not mix thoroughly immediately after they are poured into the glass bot-
tle. This prevents them from starting to react at the time they are mixed, which leads to a
systematic error in the measured reaction time that causes it to be larger than the actual
time.
Methodology: Immediately after the glucose solution is added to the diamminesilver(I) so-
lution in the glass bottle, the cap of the bottle is put on and the bottle is shaken well to
mix the solutions. This minimizes the time wasted for the solutions to mix.
Improvement: It is possible to put the bottle with the diamminesilver(I) solution in it on a
magnetic stirrer and add the glucose solution while the stirring is on. This maximises the
speed of mixing, but requires a reaction vessel larger than the 20cm3 glass bottle for the stir
bar to fit. Also, adding solutions with the stirring on may lead to splashing, which causes
errors in the concentrations of the reactants and also safety concerns.

6 Extension
Applying my findings to my initial interest of using the Tollen’s test as a demonstration of
redox reactions for children, I find that it would have the best results if the temperature is at
least 25°C, so that the mirror forms in around 1 minute.
This investigation can also be used by chemical engineers who are interested in manufacturing
real mirrors using the silver mirror reaction. It is possible to determine what temperature would
be suitable for such a industrial process that maximises the production efficiency and generates
most benefit.
In the future, it is possible to research on how the type of metal ion (Au3+ , Cu2+ ) affects the
activation energy for similar mirror reactions. These two metal ions are also known to produce
mirrors when mixed with a reducing agent through a similar nanoparticle process. It might be
meaningful to investigate how the ionic radius, ionic charge and electronegativity of the metal
affect the activation energy for reactions with similar mechanisms.
Furthermore, it is possible to measure the reflectance of the mirror in a more systematic way
than using a camera together with the video analyser. The reflectometer is a combination of my
design and the usual spectrometer that measures the spectrum of the light reflected. It is capable
of generating more precise data for all wavelengths instead of only for red light for my setup.

12
Works Cited
Benet, William E., et al. “The Mechanism of the Reaction of the Tollens Reagent”. Journal of
Chemical Research, vol. 35, no. 12, 1 Dec. 2011, pp. 675–677. doi:10.3184/174751911X13206824040536.
Breedlove, C. H. and John Softy. “The Explosive Potential of Tollens’ Reagent”. Journal of
College Science Teaching, vol. 12, no. 4, Feb. 1983, p. 281. JSTOR, jstor.org/stable/42965976.
Accessed 28 Nov. 2021.
Sigma-Aldrich. “Safety Data Sheet - Ammonium Hydroxide Solution”. Millipore Sigma, 12 July
2021. sigmaaldrich.com/HK/en/sds/FLUKA/318612. Accessed 28 Nov. 2021.
———. “Safety Data Sheet - Silver Nitrate”. Millipore Sigma, 3 Apr. 2021. sigmaaldrich.com/
HK/en/sds/aldrich/204390. Accessed 28 Nov. 2021.
———. “Safety Data Sheet - Sodium Hydroxide”. Millipore Sigma, 28 Oct. 2021. sigmaaldrich.
com/HK/en/sds/sigald/s8045. Accessed 28 Nov. 2021.
Zhang, Zude . Wuji Huaxue [Inorganic Chemistry]. 2nd ed., P of U of Sci-
ence/Technology of China, 2014.
Zhu, Aihua , et al. “Xinzhi Yiquan Yu Yinanrongye Jinxing Yinjing Fanying De Shiyan
Tanjiu [Experimental Research on the Silver
Mirror Reaction between Newly Prepared Ethanal and Tollens Reagent]”. Huaxue Jiao Yu
Xue [Chemistry Teaching And Learning], no. 9, 2018, pp. 91–93. doi:10.3969/j.
issn.1008-0546.2018.09.028.

13
Appendix

(a) Sensor data for τ = 25°C. (b) Sensor data for τ = 30°C.

(c) Sensor data for τ = 35°C. (d) Sensor data for τ = 40°C.

Figure 4: Graphs of sensor data.

14