You can generate CO2 with a mixture of yeast, sugar and water, and there is scientific evidence for its effectiveness in attracting mosquitoes. Saitoh et al. (2004) tested the apparatus shown below.
They constructed it as follows:
Two plastic bottles (2-liter volume) were used to hold water solutions of sugar and yeast. The 2 bottles contained different concentrations of dry yeast and sugar: bottle A, 150 g of sugar + 12 g of dry yeast, and water added to total a volume of 1,500 ml; in bottle B, 100 g of sugar + 6 g of dry yeast, and water added to total a volume of 1,750 ml. Because bottle A contains a larger amount of dry yeast, the output rate of CO2 is higher and the length of supply period is shorter than in bottle B. By using the 2-bottle system, we could achieve the high output rate of CO2 as well as the long supply period. The bottles were connected to each other with polypropylene tubing and to a small (500-ml-volume) plastic bottle holding the overflowed water solution. Generated CO2 was released from a 5-mm hole on the outer wall of the small bottle. For easy preparation of the water solution as well as cleaning, there were 3 joints in the connection tubing (Fig. l). The small bottle was hung close to the opening of a suction trap, similar in design to the CDC-light trap (Service 1993). It was made of 14-cm-long acrylic tubing with an inside diameter of 8.5 cm attached to a 3.0-V motor driving a three-bladed plastic fan powered by four 1.5-V dry batteries.
You can of course omit the suction trap and just hang the CO2 emitter near your existing trap, if you have one.
The abstract summarizes the results:
The system could, at average, generate 32.4 ml/min of CO2 for at least 27 h. The total weight of the CO2 generated was estimated to be 94 g. The efficacy of yeast-generated CO2 as attractant for mosquitoes was significant, and the following 6 mosquito species were collected using yeast-generated CO2 traps from July to September 2003 in a residential area of southern and northern Yokohama City, Japan: Aedes albopictus (Skuse), Armigeres subalbatus (Coquillett), Culex halifaxii Theobald, Cx. pipiens pallens Coquillett, Ochlerotatus japonicus (Theobald), and Tripteroides bambusa (Yamada). Besides mosquitoes, various other insects were collected in the trap.
Smallegange et al. (2010) used a similar yeast system and found:
Traps baited with yeast-produced CO2 caught significantly more mosquitoes than unbaited traps (up to 34 h post mixing the ingredients) and also significantly more than traps baited with industrial CO2, both in the laboratory and semi-field. Adding yeast-produced CO2 to traps baited with human odour significantly increased trap catches. In the MalariaSphere, outdoor traps baited with yeast-produced or industrial CO2 + human odour reduced house entry of mosquitoes with a human host sleeping under a bed net indoors.
Saitoh, Y., Hattori, J., Chinone, S., Nihei, N., Tsuda, Y., Kurahashi, H., & Kobayashi, M. (2004). Yeast-generated CO2 as a convenient source of carbon dioxide for adult mosquito sampling. Journal of the American Mosquito Control Association, 20(3), 261–264.
Smallegange, R. C., Schmied, W. H., van Roey, K. J., Verhulst, N. O., Spitzen, J., Mukabana, W. R., & Takken, W. (2010). Sugar-fermenting yeast as an organic source of carbon dioxide to attract the malaria mosquito Anopheles gambiae. Malaria journal, 9(1), 292.