Scientific discoveries and inventions have long been established as two distinct and sequential activities. It has nonetheless been showed that projects aiming at producing both scientific discoveries and inventions could record impressive results. Our investigations are focusing on the creativity of collaborations outputs: a first agent is entailed to design a scientific discovery and another one invention. We use fixation effects as a performance measurement indicator for creativity based on Design Theory. We propose a first set of elements that can be suffering from fixation effects in both invention and scientific models designers reasoning. We propose a series of defixed inputs that could be shared between both designers to overcome their fixation effects. We highlight that if partners are engaged in one-way knowledge transfer it can conduct to "fixation traps". We define a set of restrictive conditions that could conduct to a "cross-defixation process": both actors would be able to create conjoint new inventions and scientific models in the non-fixed design path. In particular this process does not required designers to be defixed before starting the collaboration.