Research Article | Open Access

An incubator to test the effects of LED lights on antibiotic resistance

S. Fay1* and R. M. Geller2

Author Affiliations

*Corresponding author: Sophia Fay
Laguna Blanca High School, Santa Barbara, CA 93110, USA; E-mail:

Received: September 13th, 2017; Accepted: September 26th, 2017; Published: October 2nd, 2017

Eng Press. 2017; 1(1): 24-29. doi: 10.28964/EngPress-1-104

Ⓒ 2017 Copyright by Fay S, et al. Creative Commons Attribution 4.0 International License (CC BY 4.0).


A major problem in the United States and around the world is infections acquired in healthcare environments such as hospitals and nursing homes. Furthermore, bacteria are developing resistance to the antibiotics used to treat these infections. This engineering project is part of an experiment to test if a new generation of white light-emitting diode (LED) lights‒which hospitals are beginning to use‒promotes antibiotic resistance in bacteria. This concern is based on the fact that these new LEDs contain a higher proportion of blue light (which is a component of white light) than other lighting sources, and when blue light passes through a microbe, it produces free radicals. To test the effects of LED lights on antibiotic resistance, a unique incubator is needed‒this incubator for bacteria is the engineering project discussed in this paper. The incubator has two chambers: one with regular halogen light bulbs and one with LED lights. The halogen bulb chamber serves as the control group side that can be compared with the results of the LED side. The most challenging aspect of the incubator is making sure that the temperatures in both chambers are equal. Without special engineering, the temperatures would be very different on the two sides because halogen bulbs produce much more heat than LED lights. In order to get the temperatures equal in both chambers, active control of a heater was used. Temperature changes were achieved through the use of a microcontroller that constantly sensed temperatures and controlled the heater as needed. This engineering project was a success: the resulting temperature differences obtained between the two chambers are small enough to allow the incubator to be used in the future bacterial experiment.

KEYWORDS: Light-emitting diode; LED; Incubator; Bacteria; Antibiotic resistance.


Every year, approximately 99,000 people die from preventable healthcare-associated infections, which are infections acquired in a hospital or other healthcare facility.1 Contaminated surfaces contribute to the transmission of these infections. Yellow boxes mark contaminated surfaces in Figure 1, showing that numerous surfaces have bacteria.2 Note that these surfaces are exposed to the room’s overhead lighting. A new hypothesis proposed here, is that a hospital’s overhead lighting can increase antibiotic resistance, contributing to more healthcare-associated infections. Testing this hypothesis is the overall scientific goal for which the bacterial-growth incubator in this paper is a crucial part.

A recent breakthrough has led to the development of highly efficient white light-emitting diodes (LEDs) that can replace regular lighting. Because LEDs use less energy and last longer, many businesses and hospitals are rapidly converting over to this new technology. However, despite the benefits these new LEDs provide, they can also have drawbacks. The American Medical Association (AMA) has warned that these new LEDs can disrupt humans’ circadian rhythm, which is a mild health concern.3 The key factor that makes these new LED lights a potential concern is that they contain a larger proportion of higher-energy blue wavelengths compared to halogen or fluorescent light bulbs.4 This blue light in these LEDs is behind the American Medical Association’s warning, and also our larger concern. The potential problem with a higher exposure to blue light in hospitals is based on the fact that when blue light passes through a microbe, it produces free radicals.5 Unlike humans, bacteria have no protective outer layer of pigment molecules to absorb blue light. Therefore, humans do not acquire free radicals from exposure to blue light, but bacteria do.

The free radicals within microbes can lead to genetic mutations in their DNA. With high intensities of blue light, this effect can be used to kill bacteria.6 However, an unanswered question is whether or not the fewer genetic mutations caused by lower intensities of blue light, which do not kill the bacteria, can promote antibiotic resistance. In general, there is a component of luck‒based on random genetic mutations‒for the bacteria to evade being killed by antibiotics, and a population of bacteria with a greater variety of genetic mutations has a greater chance that one of the bacteria will survive by being antibiotic resistant. In other words, a high level of mutations can kill bacteria, but a low level of mutations (possibly created by LED lighting) might increase antibiotic resistance. This split between two outcomes is very similar to the fact that high doses of X-rays are used to kill cancer cells (by causing many DNA mutations), whereas low doses of X-rays can cause cancer (by creating a few harmful DNA mutations).7

However, only an actual experiment can determine if LED lights create enough mutations to significantly increase the antibiotic resistance of a bacterial population, and no published papers on this question could be found. Specifically, the scientific question that this incubator will be used to answer is this: Do LED lights promote antibiotic resistance in bacteria? To test this idea, bacteria will be exposed to halogen bulbs in one chamber, which hospitals already use, and LED lights in another chamber, which hospitals are switching to (see overhead photo of the two-chamber incubator in Figure 2).

A test will then be performed to assess antibiotic resistance based on the “zone of inhibition” method. However, this paper focuses on the considerable engineering challenge involved in building the incubator that will be used in the bacterial experiments. The main challenge is how to keep the temperature the same in both chambers, because halogen bulbs produce much more heat than LED lights. If the temperatures are not equal in both chambers, there would be no way to tell if any subsequent difference in antibiotic resistance was due to the LED lighting, or merely the large temperature differences.

Figure 1: Hospital surfaces contaminated by bacteria, marked with yellow boxes.

Figure 2: Overhead view of two-chamber incubator-LED light chamber and halogen bulb chamber.



In order to carry out a successful experiment, all the conditions within the incubator have to be kept constant except for the one experimental stimulus (often referred to as the “independent variable”). In this case, the experimental stimulus is which type of light the bacteria are exposed to‒halogen bulbs versus LED lights (Figure 3). In the incubator, the challenging variables to keep constant are temperature and illumination. For illumination, the goal is to have the same amount of light in each chamber as perceived by the human eye, called lumens, regardless of the source or type of light. The reason to keep lumens constant for both types of lighting, is that this is precisely what hospitals seek when installing new lights: the same amount of visible illumination from a new light that also consumes less energy.

After much experimentation, it was found that the Phillips brand of lighting helped to provide equal illumination in the incubator by producing “replacement LED lights” for halogen bulbs, so that both produce similar lumens. Therefore, the halogen bulbs and their LED replacements are being used in this experiment; three Phillips 50 Watt halogen bulbs emitting 430 lumens (MR16 GU10) and three Phillips “50 Watt Equivalent” LED bulbs emitting 400 lumens (MR16 GU10) are being used, with each type of light in their own chamber. Also, to aid in uniform illumination, a sheet of halon is attached to the bottom of the cooler lid. Halon is an extremely good diffuse reflector across the entire visible spectrum which aids in providing uniform illumination on the bottom of the incubator where bacteria will be growing in Petri dishes.

A Digital Lux Meter (EPS2000LM) was used to measure the amount of lumens at various points on the bottom of the incubator in the areas where the Petri dishes sit. The manufacturer’s reported lumens could not be relied on for two main reasons: 1) The actual lumens in any one spot depend on how much the bulb spreads the light out (for example, it could be a narrow or wide beam), and the halogen and LED lights might not spread light out in the same way; and 2) Reflections within the incubator alter the amount of light landing on the bottom where the Petri dishes are. Using the Digital Lux Meter, it was found that the average lumens on the Petri dishes in the LED chamber versus halogen chamber are 1033 lm and 1095 lm respectively.


In order to ensure that the only difference between the two chambers is the type of light, it is necessary to keep both chambers at the same temperature. However, halogen bulbs give off much more heat than LED lights do, leading to a substantial difference in temperatures between the two chambers. For instance, in initial testing with no fans and no heater, the halogen bulb side would rise to 115 ºF in about 15 minutes before we turned off the lights to avoid overheating the incubator, while the LED side would remain close to room temperature at around 75 ºF. (Note: The fact that LEDs emit much less heat is precisely why they are so much more energy efficient.)

For bacterial growth, the ideal is for both sides to be at the same temperature, which is somewhere between 86 °F and 107 °F. Therefore, fans were added to cool down the hot halogen side, and a heating element was added to warm the cooler LED side. Two fans were also added to the “divider-wall” between the two chambers (Figure 4). To help keep the two chambers at the same temperature, these fans blow in opposite directions, circulating air between the two sides.

To cool down the halogen side of the incubator, two 12 volt 6 cm square fans are used. One blows in air from outside the incubator, while the other blows warm air from the inside of the incubator out, creating a steady stream of air through the halogen chamber. Several holes of various sizes were also cut through the incubator wall to vent heat. No further control on temperature is applied to the halogen chamber. With all four fans running (but no heater), the halogen side reaches a steady temperature of about 100.2 ºF after 18 minutes. On its own, this temperature on the halogen side is a good temperature for growing bacteria. However, the LED side only reaches 88.7 °F, and the incubator must keep both sides at the same temperature. To increase the temperature on the otherwise-cooler LED side, a 100 Watt, 110 volt ceramic infrared heat emitter is used.

To monitor the temperatures, an Arduino Uno microcontroller is connected to a temperature sensor in each chamber. A microcontroller is like a stripped down computer with no keyboard or screen‒it has inputs, such as the temperature sensors being used, and it has outputs, which are used to turn the heater on and off. Turning the heater on and off is performed in a special way called “feedback control”. While feedback control can be a very complex topic, only a simple application is used here. Consider trying to maintain the LED chamber at a high, but constant temperature, and call this the setpoint temperature. Whenever the Arduino measures that the actual chamber temperature is below the desired setpoint temperature, the Arduino turns on the heater. Likewise, whenever the Arduino measures that the temperature is above the setpoint temperature, the Arduino turns the heater off. A thermostat uses this same principle to maintain a somewhat constant temperature in homes. However, in order to get the two chambers to reach the same temperature, the setpoint temperature for the LED chamber is simply the actual temperature inside the halogen chamber. In other words, the heater on the LED side will be turned on and off in order to match the temperature on the halogen side.

Figure 5 shows the key features of the incubator. A rectangular-shaped cooler with interior dimensions of 17.75” L×12” W × 10” H is being used, and the figure does not show the lid, although the lid remains on during the experiment. A few things to note for the numbered figure annotations: 1. The heater for the LED chamber is mounted through the incubator wall and high-temperature double-sided tape is used as insulation; 2. The divider-wall has fans mounted that constantly circulate air between the chambers to help bring them to the same temperature; 3. Each chamber holds three 3.5 inch Petri dishes to improve the statistical significance of each test; 4. The supporting electronics are mounted on the side and will be discussed later; 5. & 8. The temperature sensors in each chamber are covered with a metal “shield” to avoid direct heating from the lights or the infrared heater on the LED side; 6. & 7. These are the light fixtures shown in Figure 3, but in the correct experimental position.

It’s worth noting that there would be more than one way to reach the goal of maintaining both chambers at the same temperature, and in the desired range between 86 °F and 107 °F. We could have used smaller external fans for cooling the halogen side (leaving in more heat) and then used larger divider-wall fans (for equilibration between the sides). We also had variable control over the heater output. Presumably, there is some ideal configuration, where temperature equilibration could be met without using feedback control, and we spent a fair amount of time trying to find that configuration. However, only by adding feedback control, could we bring the two chambers into close agreement. Furthermore, by using feedback control the incubator can accommodate experiments using a range of lighting sources, which might be needed in the future.

A Technical Safety Note

Only those considering projects involving the feedback control of 110 V devices need to read this note. In order to implement feedback control, relays had to be hooked up first. A relay is an electronic switch that can be turned on and off by the low-power Arduino, yet the relay-switch being turned on can control high-power devices like the 100 W heater which runs on 110 V. See Figure 6 for the relays and other electronics.

For the relays, there was an option to choose if the heater was ON as the default (this is called “normally closed” where the relay is “on”) or the heater was OFF as the default (which is called “normally open” or “off”). The default position of a relay is simply its switch connection when there is no power going to the relay. It is the safest to only have the heater on when the relay is being powered, meaning the heater should be connected to the “normally open” side. That way, if the power supply to the relay is interrupted, signifying a potential safety problem, the loss of power turns off the heater by reverting it to “normally open”. If it was the other way around, a loss of power to the relay would not let the heater be turned off, even if the temperature was too high and the Arduino sent a signal to turn it off.

As a final step for overall safety, a safety-relay was set up that turns everything off if the temperature on either side of the incubator goes above 105 degrees. The safety-relay connects into the main power input for the entire system, allowing everything to be turned off at once if the Arduino measures that either chamber gets too hot. As with the heater relay, the default for the safety-relay is that the main power input is “normally open,” meaning it is OFF, or not connected if there is no power. Therefore, in the loss of power to the safety-relay itself, everything will be turned off.

Figure 3: LED & Halogen Light Fixtures- Laying on Top of Incubator.

Figure 4: Divider-wall with circulation fans.

Figure 5: Overhead view of incubator with lid off.

Figure 6: Microcontroller & electronics box.


After checking all of the safety systems, the incubator was turned on to collect temperature data from both chambers. The Arduino’s “serial monitor” was used to output the temperature data. Then, the data was imported into Excel to create the plot in Figure 7. The room temperature during the experiment was 73.5 ºF, and temperatures were sampled every 10 seconds for about 18 minutes.


The data in Figure 7 show that the halogen chamber (in red) heats up more quickly than the LED chamber (in green). Recall that the temperature setpoint for the feedback control on the LED side is simply the halogen chamber temperature. This means the LED heater remained on until the LED side caught up with the halogen side, which took about 8.83 minutes, and occured at a temperature of 98.5 ºF. The data clearly settles into a steady pattern after 12 minutes, with the following characteristics. The halogen side remained at an average temperature of 100.6 ºF with a standard deviation of 0.4 ºF. Meanwhile, the feedback control turned the heater in the LED chamber on and off in trying to match this 100.6 ºF. The feedback control worked quite well, as the LED chamber maintained an average temperature of 101.3 ºF with a standard deviation of 0.6 ºF.

Figure 7: Incubator temperature plot.


It was found that the average lumens on the Petri dishes in the LED chamber was 1033 lm, and the average lumens in the halogen chamber was 1095 lm. This is only a 6% difference, and not enough to be the cause of significant differences in antibiotic resistance. Therefore, it can be considered that the two incubator chambers have equal lumen exposures.

With average temperatures of 100.6 ºF on the halogen side and 101.3 ºF on the LED side, the average temperature difference is only 0.7 degrees. There is no reason to suspect that such a small temperature difference could lead to a significant difference in antibiotic resistance, so as with illumination, the two chambers can also be considered to be at the same temperature.

To conclude, bacteria grown in this incubator will only be subjected to one difference (or independent variable), which is the intensity of blue light. Therefore, this incubator is adequate for testing if white LED lights promote antibiotic resistance.


No external funding was required for this work. I would like to thank my mentor, Dr. Robert Geller, for his help and guidance on this project.


The authors declare that they have no conflicts of interest.


1. Preventing healthcare-associated infections. Centers for disease control. Website. Accessed September 4, 2017.

2. Hospital-acquired infection. Wikipedia. Web site. Accessed September 4, 2017.

3. AMA adopts guidance to reduce harm from high intensity street lights. American medical association. Web site. Accessed September 4, 2017.

4. How is white light made with LEDs? Lighting research center. Web site. Accessed September 4, 2017.

5. Lubart R, Lipovski A, Nitzan Y, Friedmann H. A possible mechanism for the bacterial effect of visible light. Laser Ther. 2011; 20(1): 17-22. doi: 10.5978/islsm.20.17

6. St Denis TG, Dai T, Hamblin MR. Killing bacterial spores with blue light: When innate resistance meets the power of light. Photochem Photobiol. 2013; 89(1): 2-4. doi: 10.1111/j.1751-1097.2012.01233.x

7. Bushong SC. Radiologic Science for Technologists. 8th ed. St. Louis, MO, USA: Elsevier Mosby; 2004: 482-548.

Volume 1, Issue 1
October 2017
Pages 24-29

All for Joomla All for Webmasters