SecSpaces is a data-driven coordination model that supports a tuple-based coordination space extended with mechanisms for controlling and authenticating the access to entries. This is achieved exploiting the notion of (symmetric and asymmetric) abstract partitions inside the space. In this paper we consider one of the limitations of SecSpaces: it is not well suited for supporting the atomic access to more than one partition at once. In order to tackle this limitation we extend the SecSpaces model introducing an operator to combine partitions; output operations can use the new operator to make an entry visible in more than one partition, and data-retrieval operations can use it to access atomically more than one partition. We formally define, in terms of a process calculus, this notion and we demonstrate the flexibility of this new mechanism via examples