Résumé : Classical cryptography offers solutions for various applications like distributing keys, encryption, authentication, etc through classical algorithmic operations. Consequently, they can be attacked by similar methods. Quantum cryptography on the other hand uses quantum states to encode information. This also means that the information is protected by the intrinsic properties of these quantum states, making attacks more difficult to design and generally detectable or even impossible. The most famous quantum protocol, quantum key distribution, was published in 1984, but in recent years the application domain of quantum cryptography is enlarged and now it covers a very large range. One sub-domainexplores Quantum Key Recycling and this is also the research domain in which this thesis is located.In Quantum Key Recycling, the aim is to re-use (a part of) the shared secret keys that are used in quantum encryption and quantum authentication schemes. Since the ciphertext and/or the authentication tag is a quantum state, the detection of eavesdropping is made possible. In case no eavesdropping was detected, the secret shared key can be safely re-used which is impossible for classical information-theoretic encryption and authentication schemes like the one-time pad or Wegman-Carter authentication.In this thesis, we first design a secure quantum key recycling scheme that is within reach of being implemented with current technology and that allows to re-use the shared secret key entirely. Our main objective was to minimize the quantum communication between the participants. Moreover, the keylength is reduced compared to other schemes that address the same problem, more precisely encrypt and authenticate classical messages without the need for a quantum computer. The security proof of the scheme is based on guessing probabilities and trace distances. With this scheme, the authentication problem between two mutually trustworthy participants is answered. Therefore, we also considered quantum authentication with arbitration among two distrustful parties. Despite the fact that several quantum authentication schemes with arbitration were proposed in the literature, we can find security flaws with respect to their requirements, i.e., verifiability, unforgeability, non-repudiation and resolution for disputes. In order to solve these security problems and thus meet all the requirements, we propose and analyze anew quantum scheme. The scheme does not use entangled states or complicated transformations in contrast to several other schemes. Moreover, we analyze the security of our scheme in a formal way, which is also lacking in most publications on arbitrated quantum schemes.