This paper investigates plasmons in doped graphene at infrared frequencies, focusing on their low losses and wave localization properties. The authors show that for frequencies below the optical phonon branch (ħω_Oph ≈ 0.2 eV), plasmons in doped graphene can exhibit both low losses and significant wave localization. Large plasmon losses occur in the interband regime, which can be pushed to higher frequencies with increased doping. For sufficiently large doping, there is a frequency range from ω_Oph to the interband threshold where plasmon decay via optical phonon emission is inactive. The losses are calculated using random-phase approximation and number-conserving relaxation-time approximation. The measured DC relaxation-time is used as an input parameter for impurity collisions, while optical phonon contributions are estimated from electron-phonon coupling effects on optical conductivity. Graphene, a two-dimensional material with unique properties, is a promising candidate for plasmonic applications due to its ability to support low-loss plasmons in certain frequency ranges. The paper compares graphene plasmons to surface plasmons on dielectric-metal interfaces, noting that graphene plasmons can have better wave localization and lower losses. The study shows that for frequencies below the optical phonon frequency, plasmons in doped graphene can have significant wave localization and low losses, making them suitable for nanophotonic applications. The results suggest that graphene plasmons could be useful for nanophotonic devices, especially in the infrared range. The paper also discusses the importance of considering electron-phonon coupling and impurity scattering in calculating plasmon losses. The authors conclude that further research is needed to refine the understanding of plasmon relaxation times and their dependence on various factors.This paper investigates plasmons in doped graphene at infrared frequencies, focusing on their low losses and wave localization properties. The authors show that for frequencies below the optical phonon branch (ħω_Oph ≈ 0.2 eV), plasmons in doped graphene can exhibit both low losses and significant wave localization. Large plasmon losses occur in the interband regime, which can be pushed to higher frequencies with increased doping. For sufficiently large doping, there is a frequency range from ω_Oph to the interband threshold where plasmon decay via optical phonon emission is inactive. The losses are calculated using random-phase approximation and number-conserving relaxation-time approximation. The measured DC relaxation-time is used as an input parameter for impurity collisions, while optical phonon contributions are estimated from electron-phonon coupling effects on optical conductivity. Graphene, a two-dimensional material with unique properties, is a promising candidate for plasmonic applications due to its ability to support low-loss plasmons in certain frequency ranges. The paper compares graphene plasmons to surface plasmons on dielectric-metal interfaces, noting that graphene plasmons can have better wave localization and lower losses. The study shows that for frequencies below the optical phonon frequency, plasmons in doped graphene can have significant wave localization and low losses, making them suitable for nanophotonic applications. The results suggest that graphene plasmons could be useful for nanophotonic devices, especially in the infrared range. The paper also discusses the importance of considering electron-phonon coupling and impurity scattering in calculating plasmon losses. The authors conclude that further research is needed to refine the understanding of plasmon relaxation times and their dependence on various factors.